The two situations are not comparable.

One is the "production" of music and the other is the "reproduction" of music. The former is purely subjective while the later is, for the most part, not subjecticve at all.

A musical instrument speaker usually has a lot of harmonic distortion because this makes the instrument sound richer. The player may even add more distortion. This is the artists right. But I have no philisophical right to add anything to the performance if I want to call it reproduction.

And this is a very good question:

"So why can't that same position be applied to audio reproducing that Stradivarious?"

If listeners all listened to the Strad and then lsitened to a reproduction as a direct comparison then you have the beginings of a valid controlled listening test. (You also have to control the environment, playback level, lots of things). But then the Strad doesn't really make any difference does it? I mean I could just as easily use a really crappy violin and as long as its poor sound was reproduced then the test is just as valid.

Now what if the sound system some how made the poor violin sound good, perhaps even better than the Strad. Would this then be a "good" sound system?

You see you have highlighted a strong point that I will make in the text. That someone can only evaluate a sound system for good reproduction if they are intimately familiar with what is being reproduced.

A classic example is having a musician listen to two sound systems with different conductors and orchestras playing on each but playing the same piece. They will almost inherently pick the one with the better performance as sounding better. It's easy for musicians to hear the notes, but it's much more difficult for them to hear the sound. You see the two things are quite different - a point that is often missed.

I have a good friend who is a well known concert pianist. He's teaching me to hear the "performance" and I am teaching him to hear the "sound reproduction".