I have a hypothesis on this subject. Right now, it's still sort of in the developmental phase, but I think I'm on the right track. Anywho...
I believe that a lot of what we call "the piezo quack" is actually the sound of an impedance mismatch between the guitar and the rest of the signal chain. I also believe that the piezo pickup in an ac/elect. guitar is particularly sensitive to impedance loading effects.
I have a cheap ac/elect (an Ibanez that retails new around $500) that I use with some of the acoustic gigs that my band plays from time to time. Right away, I noticed that it seemed to be uncommonly sensitive to the type of amplification that you plug it into. I have one bass rig that is so flat and neutral that it is, essentially, a full range PA system. I assumed that the Ibanez would sound best through that. WRONG! It sounded terrible and the piezo nastiness was particulary strident. This tonal characteristic changed enormously, depending upon what I plugged it into. I got the best results with that particular guitar by plugging it into a Traynor guitar amp. (Please note: it's an electric guitar amp! Who'd have thunk it?) It sounds remarkably good. No sign of a piezo quack anywhere! Since the gigs that would require me to play that guitar are all very small, intimate acoustic gigs, I simply play it through that amp. If I DID need to put it into the PA though, I'd just take the XLR out of the back of that amplifier.
Different guitars yield very different results, though. Indeed, I've heard of pros who plug their ac/elect guitars into various kinds of stomp boxes (eq, distortion, whatever) -- but without actually switching the pedal ON -- because it supposedly makes it sound better. This can only be an impedance matching effect. Now that my band has secured a drummer and returned to playing amplified electric gigs, the only acoustic that we are using on stage is a Martin with a piezo/mic system onboard. The mic feeds back too much, so we use the piezo exclusively. This guitar actually sounds pretty darned good plugged direct into the board. I also had him experiment with various stomp boxes in the signal chain, but it didn't seem to help much. The best results that we've gotten so far are to put it through a cheap Whirlwind direct box, and then into the board. Sounds really good.
My advice is to experiment with various devices to see what makes your guitar sound best live. Don't be afraid to try stuff that seems unconventional. I believe that the most critical thing is to find what will make your guitar match up best to the impedance of the board (or whatever you choose to plug it into).
Anyway, that's my 2 cents.