I'm not so sure there isn't some way to scientifically ascertain and record subjective impressions and thereby preferences, but I agree it hasn't been done.
Read my article on controlled testing Hedonic analysis is well established. Floyd Toole wrote a whole book about it.