I've done a lot of work in audio and I agree completely with this article. The only true way to tell if someone is or is not hearing a difference is through an ABX, where you play somebody two known sources (A and B) then an unknown source (X) and the user must identify whether X was either A or B. There was a post on the consumerist a while back where some guys did a comparison of Monster speaker wires and coat hangers, none could tell the difference.