And where is your proof that it doesn't exist... Unless you've pregraphed 2 pairs and noticed no difference, along with some brain scans to prove it's psychological, you can't say much either...
If it was truly psychological, mental, or placebo, everyone who heard a difference would hear differences at a different rate, and different times as each person's psyche will adjust differently. However, several outcomes show that we all hear the changes, close to uniformly.
Also, ABing them won't prove sufficient because you'll psychologically want to not hear differences... Quite honestly, that argument is trite, overused, and works for both sides...
Also note, so far, every test done regarding burn in has shown some sort of actual change in sound (although all may not be audible). All tests (yes, this uses FR graphs, data, ect.) show some sign of burn in... You're up at bat now. Show me 2 graphs that show no signs of changing whatsoever (Tyll's graphs show change that is outside the error rate showing that those do change, they just change really slow; 65 hours may not have been sufficient enough for them to burn in).