Gugliemo,
I can understand your philosophy, and the intent on which you base it.
But there's another point to consider in the pursuit of live sound, and the tools we have to work with.
Unless deliberately manipulated by us, a synthesiser is capable of producing a tone at precisely A440 if that is our parameter. In a visual/orchestral sense this would be the equivalent of placing a microthin card on the string of a violin and bowing that string with a device capable of a perfect rate of pressure and friction for a period of time. Even then we would not use the start of the sound, because the initial push of the bow would affect the tuning as the vibration of the string begins its 'journey'. No vibrato, and absolutely no deviation in the width of the string oscilation might get us close to that A440 parameter.
The reality is, the bow is operated by a human being, incapable of a perfect stroke just by the nature of our ' human machinery'. Likewise, the pitch is determined not by the microthin card, but the width of a finger. No two human beings have fingers the same width, so each human equivalent of A440 is going to be different. Then introduce expressive content like vibrato, and A440 goes from being microthin and precise to a broad sound that is, again in reality, more accurately described as A440 +/-, and then some.
Additional factors?
At the start of each note in a 1st violin section, no matter how good, each player instinctively 'finds' the group tuned note. So a semibreve played adagio is more in tune after the start of the note not at its precise introduction. Then the player is influenced by the note to follow. In the key of c, a section playing a b natural will, if written as a final cadence, be 'instinctively conscious' of the B natural's role as a leading note, and its been my experience, this note will be a little sharper in pitch, and even further away from 'A440.'
We have been conditioned by years of artifically produced sounds, and our attempts to manipulate them, into believing that a collection of musicians playing live is 'A440' capable, and nothing could be further from the truth. That live sound we strive for is not the product of singularly fine tuning, but the rare and wonderous quality of humanity, and its incapacity to achieve 'perfection.' Instead we gain a new perspective of what is ideal from expressive and accurate playing, the quality of the composition and orchestration, and the 'passion' and 'humanity of the performance itself.
VSL have produced a library that in itself, brings us closer than ever, and better than any other product, to that ideal. Not because each note is 'A440' perfect, or that every bow stroke is mechanically perfect, but for the humanity in the samples, those tiny differences between each note, dynamic, and bowstroke. They've gone as close as i think they should to the 'ideal', and any attempt to fine tune those samples to the nth degree you desire will certainly remove the 'humanity', and take us further away from our collective compositional aspiration. VSL is certainly accurate, more so than others, and the precision is wonderful, but best of all its still human, something other equivalently marketed sample libraries singularly fail to achieve.
I've sat in orchestras and played, and you'd be surprised how broad tuning is, how much it varies according to harmony and cadence and individual playing techinique, and how all these and other inaccuracies add up to a sound that is human and brilliant. Instead of using a note that for you is 'out of tune', why not use a short note of another type? Will this new note be so far removed from the intent of your phrase, to be unuseable? Or will it give you the opposite? a 'human' sound, subtly different and yet still well within the broad brush of your piece?
Regards,
Alex.