Good points about kinesiology and especially the idea that "generic solutions only go so far" when customizing an instrument to a person's particular abilities.
I would be very interested in any ideas for transpositionally invariant fingerings. I started exploring this because of similar work done with the EWI (especially by Bret Pimentel and Curt Sipe).
I also have code I am happy to share (Mathematica / Wolfram Language notebooks).