"Singers and players of non-fixed instruments tend to instinctively go for more pleasing intervals when not constrained by the accompaniment of a fixed, equal tempered instrument on which most intervals are actually a compromise. "

Interesting thought - but it seems inherently contradictory to refer to the idea of comfortable intervals unconnected to fixed points in support of the notion that B# and C are somehow seperate identifiable points on the spectrum of sound.

How can a notion of arbitrary tonality be used to support a notion of fixed tonality?