it Does make a difference, and that difference can be huge,
If you use the wrong Type and of course ,the wrong value.
Use Ceramic- (here's the kicker-) class 2, tolerance +/-20%
value .022 (darker-.047) Don't sub the class or tolerance.
Don't care if you don't believe me, but I have done exhaustive
testing, its interesting to me to note the many parts used in the
Early days -because the were cheap -Are still the best to use
today -even though we have so many choices. As R.G. would
say -"A happy accident" There are 
exceptions, of course, But this isn't one of them.

The recommendation (although it's more like permission not to
feel guilty) for using a ceramic disc is fine by me. Other types
may be more appropriate for FX or hi-fi design, but ceramic
on an axe is hunky dory. I'm sure the Gerald Webers and Dan
Torres's of the world might have some arcane notions about
Sprague Orange drops or something, but I doubt whether
other types make any difference once you've gone unbuffered
through a budget 25' cable into four 12's full tilt.

The *value* of that cap is something that long-time
AMPAGE'rs know I have a touchy spot for. (Cripes, I'm
starting to break out in a Billy Idol-cum-Elvis twitchy sneer
right now!). When it comes to tone cap values, I think our
"happy accident" from the 50's has turned into a blind
juggernaut.

You can lash me to the mast and beat me senseless (in fact,
please do!...oops, wrong forum), but I think there is no such
thing as one-tone-cap-fits-all. In fact, I firmly believe that front
and rear pickups should have *different* tone caps. 

The industry's steadfast use of one value, regardless of pickup
voicing or pickup use strikes me as just plain boneheaded.
Because of the different harmonics emphasized by the location
of each pickup, the optimal turnover frequency for its tone
control is different. When it comes to rear pickups, what the
user wants in a guitar-mounted tone control is something that
adjusts from sharp to round. When it comes to neck pickups,
the user wants something that goes from vocal to mute. If I
wanted mute, I *wouldn't* be using my rear pickup. Cap
values of .022 to .047 may work okay for neck pickups, but
are as useless as bull mammaries for rear pickups. I would
recommend something from 6800pf to .018 (ceramic disk, of
course) in its place, ESPECIALLY if you are using a single
coil sized pickup. 

If you have a regular Strat (or clone) with dual tone controls,
take it apart RIGHT NOW. That's okay, I'll wait.
Doo-dee-doo-dee-doo. Dum-dee-dee-dum....okay, found
the screwdriver? Good, now take the pickguard off and you
will see that in the interests of economy, Fender or whomever
has seen fit to stick ONE capacitor in there to service both
tone controls. Take your soldering iron and make that cap
serve the middle tone control ONLY. Now, for the bridge
tone control, solder a 6800pf-.015uf cap in there (depending
on your tastes and tone pot value), and you will be much
happier. End of story.

When it comes to double-coil pickups, a cap of .01-.02 may
be more appropriate for the bridge pickup tone control
(different turnover point again). Going higher than a .02 just
means that you'll never turn your tone control down past 7. 

Bear in mind that tone controls normally have a *progressive*
effect or "slope". Typically, the slope is 3db/octave. So, with
the control turned down full (at which point the tone control is
at 0 ohms, so that you effectively have a signal going straight to
ground through a cap), an octave above the turnover
frequency set by the cap value, the signal is down 3db, an
octave above that, down 6db, and so on. 

Suppose I have a "turnover frequency" of 1000hz. At full "off"
rotation, there is a level drop of 3db at 1000hz, relative to
500hz (because that is one octave below 1000hz), a level
drop of 6db at 2000hz, 9db at 4000hz, etc. 

Three DB is essentially "twice as much", but twice as much of
what? If the control is only turned down a smidgen, then it's
twice as much of a smidgen, and twice as much again one
octave further out. If I start the cut lower down in the
frequency spectrum, then even a smidgen starts to snowball
into something more substantial several octaves up, once we
start doubling. If I start higher up, but turn the tone control
*down* more, then I may have a higher turnover frequency
(starting point), but "twice as much" starts to have a noticeable
effect closer to the turnover frequency.

For my tastes, larger tone caps tend to have a treble reducing
effect across too much of the frequency spectrum for bridge
pickups. Even when you turn them partly down, you've
reduced the upper midrange a bit, in addition to the treble.
With smaller caps, you can still reduce the treble, but the
midrange tends not to be affected so much, because the
turnover frequency is higher. So, you retain the personality of a
bridge pickup, but can adjust the edginess of it. As you notice,
more of the control's range "feels" usable because you can still
keep the guts of the pickup's sound even when you turn down
the tone more.

If I may use a physical metaphor, consider going from the top
of the Grand Canyon to the bottom. Now, you can leap from
the top, OR you can take a long gently sloping path. The
actual DESCENT is the same in terms of height, but the
experience is VERY different. Similarly, the amount of cut at
higher frequencies may be the same with different
combinations of control rotation and cap value, but what
happens on the way to those higher frequencies (i.e., the effect
on midrange) is different.

Incidentally, do not confuse the effect of a resistor and cap
going to ground with the effect of a resistor in the signal path
and a cap going to ground. For the first one, varying the
resistance influences how much treble loss there will be at the
same turnover point. In the second case, varying the pot
leaves the slope intact, but CHANGES THE TURNOVER
FREQUENCY. So, you would have a treble cut of
3db/octave, but it may start anywhere from 500hz or so up to
10khz, if you felt like it. 

If you go to any of the schematic archive sites mentioned here
(RG Keen's, Aron Nelson's, Jack Orman's, etc.) you may see
a posted schematic for the Proco Rat or the Bixonic
Expandora distortion units. Close to the "exit" of the effect,
you will see the tone control consisting of a 100k pot, in series
with a resistor (which helps to set the lowest possible turnover
frequency), and followed by a capacitor going to ground. This
is exactly the filter I just alluded to, and the reduction of treble
it provides is much greater than what you get on guitar tone
controls, even though both use a pot and cap.

Finally, turning one's tone control up all the way does not
mean that there is *no* treble loss. There is still a path to
ground through the control pot and cap, and the twice as much
of a smidgen rule still applies. Some players prefer to
disengage the tone control from the circuit altogether so that
there is no path to ground through the tone cap, and one does
tend to notice a bit more sheen and sparkle in the sound. Of
course, if you have humbuckers and play through 12"
speakers with the treble turned down, you may not notice the
high-treble loss through the tone control. On my own guitar, I
just use a 3-position toggle for tone. Middle is no tone cut
(i.e., no connection through a cap to ground) and the other
two positions are two different cap values yielding full
3db/octave slopes at two different turnover frequencies. I find
it perfectly adequate. 

I think what I was trying to bring out in my own post,
is that the "tolerance" of the ceramic cap is Most Important,
not the value. the tolerance of most ceramic caps 
is an almost unbelieveable +80 to -20 percent- at this rate 
the value chosen is almost a moot point- (see what I mean?)
The tolerance of the better ceramics is + or - 20 percent,
so then the "value" chosen would at least be close.
On a Strat (with no other tricks) I prefer a .022 on the 
neck PU and a .01 or a .022 (which measures on the low
side-like .015 or so.)on the bridge PU , with no cap on the
middle PU. since its use is mainly for quack in position #2 and
#4. I also like to lower the middle PU , which helps fight
"stratitis" , gives great quack ,and promotes sustain by reduced
magnet attraction- Try it. Also something to think about- The
way you mentioned doing the caps and most 
other methods that are used, end up putting the caps in parallel
on one or more of the pickup selections- This is not smart, we
all know that caps in parallel Add their value.

As you allude to, it is important to realize that WHERE the
tone control is in the circuit is important. Ideally, tone controls
should PRECEDE the volume control. Where there are
separate tone and volume controls, either the tone controls or
both tone and volume controls should precede the pickup
selector switch, in order to keep them from interacting or
having summing effects. E.g. two tone controls w/.02uf caps
turned all the way down, would be equivalent to sticking a
.04uf cap on the output, if the tone controls followed the
volume control.

As an aside, this always struck me as a VERY big downside
of guitars that had their switching system at the complete
opposite end of the instrument relative to the controls, like Les
Pauls. How does one run cables from the pickups to the
controls, then across to the switch and back to the output
jack? Even if you did, how do you do so without incurring
noise and tone sucking from cable capacitance?

As for the tolerance vs value debate, my own experience with
ceramic discs is that they are generally pretty close to what
they say they are, unless heat or something equally extreme
comes into play. The +80/-20 spec refers to off the shelf
value, not an operating fluctuation, and is largely like a
contractual loophole ("Yeah I know it SAYS .047uf, but
.059uf is still within the terms of our agreement so get off my
case"). If anything, even with an 80/20 tolerance spec, modern
fabrication techniques should probably achieve more accurate
values on a fairly consistent basis. I'm constantly amazed at
how easy it is to match caps withing 5% for filters and such
with my meter. Rather than look for types of caps, just
measure them and use the value that is appropriate for your
tonal goals. If it *says* .01uf, and *measures* .013, and it
works for you, fine. 

"...have you tried using any linear tapers for tone control pots?
Might there be any advantage in useful control range over the
usual audio or log taper pots?"

I honestly couldn't tell you what I have and haven't tried. Hell,
I'm not even sure if the volume pots in my guitar are even
audio taper. I just go through the box of pots until I find
something of the desired value that tests good and use that.

Having said that, I can see arguments offered for all kinds of
tapers, from the anti-log through linear through log or
combination tapers. It really depends what you are trying to
do and what you are trying to do it to.

For example, the late Danny Gatton (listen and be humbled)
had a tone control on his Tele that could be used like a low Q
wah. The effect was contingent on having a modest amount of
pinky action produce a robust effect. Somewhere in there is a
combination of specific taper and cap value, such that the
"voice" of the guitar would be introduced easily, otherwise it
would hamper picking style. For all I know, the remaining 180
degrees of rotation might have produced negligible change.

Other players might view their tone pot as a "mood changer",
such that the ideal pot would be one that easily identifies and
distinguishes a few different tonal moods (strident, confident,
sombre) but doesn't have to do a heckuva lot otherwise. Still
others want the tone pot to help achieve sonic balance,
perhaps by compensating for the room, the speakers, or FX
that add harmonics.

So much for the "mission" of the tone pot. The context in
which one tries to accomplish that mission is another thing.
Different pickups and guitars may have different resonances,
and the point in the pot's rotation where that resonance is
affected (resulting in the perception of a change in the tonal
"character") may vary. This is, of course, the source of my
unswerving crusade about different cap values for different
pickups. If the player routinely uses a distortion device, then
different points in the rotation may be "magic" with respect to
changing the balance of harmonics.

Personally, I would vote for the following general protocol in
nailing down what you consider to be the "ideal" tone pot.

1) Start with a larger value pot, say 500k-1meg. Doesn't really
matter what taper, for the time being, although linear might be
better for doing your calculations. High values are good for
being able to fake NO tone pot (i.e., conserve all possible high
end in the signal). Some folks like 250k volume pots for single
coil pickups to take away the brittle-sounding high end, but
even with that compensation, a 500k tone pot will let you
keep ALL of the tone you like.

2) Select a cap value in the ballpark; .02-.047 for front
pickup, .0082-.015 for rear or single coil. If you aren't sure,
just stick an assortment of caps across the two outside lugs of
the volume control and think long and hard about whether the
voicing provided with the cap is something you would like to
play *with* or something that you would rather get around.

3) Wire up the tone pot as per usual, with the preferred cap,
and turn the tone up full. With the amp's treble up full or just a
bit past where you would usually position it, rotate the tone
pot bit by bit until you start to hear what you think of as a
taming of treble. Measure the pot value and write it down.
Turn the pot a bit more and note the point where the vocal
aspects of the pickup start to be affected (i.e., upper bass,
lower midrange starts to be affected).

4) Now you can plan out the taper that works for you. 
- what is the difference in resistance between max and the first
change?
- where would you like that changeover to be situated on the
control; between 9 and 10? between 7 and 10?
- what is the difference between max and the second change?
between the first change and the second? where would you
like that one to be?

So, suppose that your ideal first changeover comes at around
the 300k mark, and you'd like it to be situated at around the
"7" mark on your tone control. So, we intend to have 30% of
rotation (number-indexed knobs usually can be treated as
referring to percent of rotation) accounting for 40% of the
resistance change. Perhaps your next changeover point comes
at around 50k, and you'd like it to happen at around the 4
mark on the knob. What happens after that is just more
muting.

Okay, so now we have the following requirements:
- turn down about 30% to get 40% resistance change
- turn down about 30% more to get an additional 50%
resistance change

Plot this out, then go to RG Keen's site and look at the
document on pots Then figure out which taper comes closest
to what you want, and - using the tips listed - mod the pot
from there.

http://www.geofex.com/Article_Folders/potsecrets/potscret.htm

Just bear in mind, that the document at GEOFEX is largely
pitched at using pots in voltage-divider mode. You will want to
use the pots as variable resistors, which will mean bridging one
of the outside lugs and wiper, in addition to use of a parallel
resistor. 

(reply)
2. Put a tone control on the bridge pickup by moving
one wire on the selector switch. Move the wire that
connects the middle pickup to the tone so it connects
the bridge pickup to tone instead. Also real easy!
 
3. Change the tone cap to .015 or .02. This leaves more
mids in the guitar signal when you roll the treble off.
Gives you a squawkier tone that's not so woofy when
you roll off the treble.