Hmm... I'm familiar with this having played first and second violin in at least one very good orchestra as well. The idea being that the conductor is there to alter the performance as he sees fit. I remember us being told not to rush, or to play other passages in a certain way. However, the whole reason for doing this was that we were going to put forth a live performance, where we had only one (per performance) chance to play it as close to as directed.
However, what I'm proposing is slightly different. Think of it as instead of a theatrical performance, it's a movie, where people can do as many takes as necessary and you can actually do each take slightly different and the director can put it together differently later. (Or even worse, alter your performance with CG! heh) In this way, you could perform once and screw up, and then submit a better performance (or partial performance) later. It would be a continual iterative process, and thus the Open Source analogy.
I think we sort of already do this though, especially with those boy bands who can't sing; they get auto-tuned until it sounds in tune. You can get better music with less skill. As awful as that sounds... 8) Maybe it's better to say that we can extend the skills of even the best of us and allow people to choose.
By the way, in terms of remote conducting, they did already try this with a Russian conductor in Vancouver conducting an orchestra in Moscow Red Square in real-time during the closing ceremony of the Vancouver Olympics. http://mariinsky.rt.com/news/gergiev-orchestra.html I don't think it worked out just 'okay' but it would be interesting to explore how technology can continue to influence something like classical music.
Anyways, thanks for the discussion from everyone. This has been very fascinating.