Mark, I think one tells the difference by noticing a family similarity in certain overtones on nitro-finished guitars among various brands. Just as you can usually tell the tone of a humbucker from a single coil due to their being part of different "tone families."
I definitely hear it, but I haven't got command of the right words to describe it correctly.
I wish I did.
Meantime I have no idea whatsoever if my theory is right, but it seems like it'd be worth testing.
Les, this is actually a lead in the direction of how you would go about testing a hypothesis such as "Do guitars with a nitrocellulose finish sound noticeably different from guitars with an acrylic/polyester/whatever finish?"
For whatever reason, it seems like the vast majority of people on these threads assume the only way to do that is to compare two identical-but-for-the-finish guitars. Well of course we all know that two guitars are NEVER "identical-but-for-the-finish." It's a nice idea, but it's totally impractical as a scientific method because it simply cannot be done. Even in the case Markie described, (A) you can't assume that the steps involved in scraping off a finish and applying a new finish didn't contribute something that has nothing to do with the finish material, and (B) you couldn't make any judgements based on the sound because the listening windows would be too far apart in time to yield any meaningful analysis.
There is another scientifically and statistically valid way to test for differences, though: with groups of guitars. (and groups of players/listeners) Sure, there are more variables, but as long as some reasonable effort is made to control for obvious ones (different body shapes, different types of pickups, etc, etc) the subtle differences that always exist between guitars with "the same" woods, shape, & pickup type and configuration, will end up coming out in the wash as normally-distributed error terms. As long as your
Ns are large enough, you can test for statistically significant differences. Not an easy or inexpensive study to undertake (who's gonna pay for this?!?) but at least it's feasible.
In the absence of such studies, I ask myself, "who plays and hears a large enough number of guitars to have a reliable judgment as to the sound of different finishes?" One answer to that is (or should be) pretty freaking obvious: Paul Reed Smith, hisownself.