Very interesting thus far, and a couple of folks are close to the real answer. I spent over 25 years as a symphony musician in a major orchestra, and then upon retirement, another 30 years as a recording engineer. I hope these credentials might add some weight to my comments.
If you stand next to a single violinist while they are playing, you can distinctly hear the horsehair of the bow and the rosin that is put on the bow making a totally non-musical noise. It is the mechanics required to make any bowed instrument play that you hear regardless of how beautiful a tone the performer manages to extract from the instrument once that bow brings the strings into vibration. The rich sound that we all crave to hear does not mostly emanate from the top of the violin. The top of the violin is where most of the mechanical noise comes from because that is where the bow and the strings intersect. The richest sound of the violin is given off by the back of the instrument. On many occasions, I would place a microphone below and to the rear of a solo violinist to capture this rich sound. A big mistake is to place microphones above and too close to the string section. That most certainly results in capturing too much of the afore-mentioned mechanical noise. Another factor is that not only the different timbres of each individual violinist, but the fact that they are almost never playing the same pitch. I'm relating this to pitch alone, not vibrato, which itself is a rapid changing of pitch. Think of it : Do you really think that the concertmaster in the front of the first violin section can hear what the player in the rear of the second violins is playing ? Not only must they sit at a reasonable distance from each other to have enough bowing room, they must forego the luxury of sitting near to each other as the woodwind and brass sections can easily do. Much of the "hash" that happens is micro differences in pitch caused by that very necessary seating arrangement. The subtle differences in timbre of each instrument is also a consideration. Woodwinds and brass also have an additional luxury, and that is that there is usually only one player on a part, not massed players like the strings.
I once was asked by the concertmaster of my orchestra, why his home recordings of his playing didn't sound as rich as he thought he sounded to himself. I pointed out to him that he was hearing a far more complex sound than anyone else possibly could because he had his instrument firmly clamped between his collarbone and his jaw which provided a "mainlining" effect directly to his hearing apparatus. He heard overtones that no one, even someone standing beside him, could hear.
It is truly remarkable that massed strings sound as good as they do when performed by a professional orchestra when considering all of the possible ways for this delicate balance to go awry.
I recorded a lot of choral groups in my career and will guarantee you that any two sopranos in a group of ten who are singing even slightly out of tune with each other will produce the "hash" we all hate. In my early days, I thought it was "phase distortion", where the signal was reaching different microphones at different times, but even by using a minimal number of microphones and keeping them at a respectable distance, the phenomenon persisted. Again, as in a violin section, the more sopranos involved, the less able they would be to actually hear each other, compounding the problem.
So, physics is involved, mechanics is involved and artistry is involved.
Just like you wouldn't want to remove the noise of Buddy Rich's sticks hitting his drumheads, you wouldn't want to remove the initiation of the string sound. They aren't synthesizers, you know ?
Definitely a fascinating subject !
