Do you believe in Magic?

Mrtennis wrote,

"another example is claiming to hear three guitars, and consulting the liner notes, finding out that there are only two guitarists.

there is an important lesson here. one should not consider the results of listening as knowledge , as sense perception is inherently unreliable."

Your example could simply be a case of incorrect liner notes and/or overdubbing.  Difficult to say but it might not have anything to do with perception being unreliable.  At the end of the Jimi Hendrix track, Axis Bold as Love, there is a long guitar solo that sounds like three guitars, but is actually only Jimi's guitar (which may or may not have been overdubbed).

Mrtennis wrote,

"i suppose its a philosophical issue, namely that knowledge can not be derived from sense perception."

I prefer to switch your statement around to read, all of our knowledge is derived from our sensory perception. If you subscribe to the evolution of man, and it won't bother me if you don't, man's sensory perception was developed primarily as a survival mechanism - to be able to see and hear an approaching predator. Since predators had excellent senses of smell, vision and hearing, man's senses needed to be very keen. Early man would have had to trust his senses completely since it was literally a life or death situation. He would have had to be able to hear a twig snap at fifty yards, differentiate a harmless animal from a tiger at two hundred yards, etc.

"it is very easy to make a mistake and misidentify an instrument. there are many examples of recordings where that is so."

I think much of the problem you refer to has to do with the playback system, many of which are unable to reproduce the instruments with their proper color, timbre, pitch, body, dynamics, etc.

"certainly our jury system provides many examples of the unreliability of witness accounts."

But witnesses are sometimes unreliable because adrenaline clouds the analytical mind, long term memory issues, or because witnesses were simply not paying attention fully. In audio these issues don't exist, at least to any great extent, while one sits in his easy chair in front of the speakers and contemplates the sound.

"it's obvious that perception is unreliable. i can cite many other examples. i doubt anyone would allege perception is perfect."

I don't think anyone would argue perception is perfect; on the other hand I am not convinced that our sensory perception is as flawed as you seem to be suggesting. I think the issue might be developing and trusting our senses.

"again , in the field of psychophsics, the just noticeable difference for distances, loudness and frequency identification empirically demonstrates that our senses lack precision."

On the other hand, you could also make the statement that it's the precision of our senses that allow us to hear various types of distortion, or the lack thereof, glare, grain, transparency, height of soundstage, air around instruments, and the difference between a Strad and a Guarneri.