Several of this points are known, thought by Peter, myself and others who frequents this forum.
http://www.physics.sc.edu/kunchur/Acoustics-papers.htmThere are discussions open on the asylum, stereophile and hydrogenaudio forums (i stay away from the latter 2)
http://www.audioasylum.com/cgi/t.mpl?f=prophead&m=54733A couple of notable points:
"There are no periodic signals in music. Everything has a starting and stopping point, and a lot of it is transient behavior. This hits at the very flaw of Nyquist/Shannon theorem."
Multitone signals are already 10 times better (to measure THD for instance), but lots of work has to be done in this subject.
I also know a person who confutes Fourier theoreme, because of time variancy (with digital sources).