Yes, ABX is indeed a wonderful thing. It's staggering the degree to which psychological bias, in the absence of blind comparisons, persuades us to believe we're hearing certain qualities. When ABX protocols (or similar scientific controls) are introduced, these differences evaporate.
That said, there are reasons why some audible differences occur with fairly casual listening comparisons. If you are comparing two receivers, and send the signals through the preamp/tone control sections, the latter will often introduce slight emphasis or de-emphasis right in the midrange, where the boost/cut section of the controls "hinge," which is exactly where our hearing is most acute.
These preamp sections are never perfectly flat, which may explain why some find the H/K units "warmer" than some other brands.
The only true test of the amp sections of receivers or integrated amplifiers would be to use the "amp-in" jacks on both receivers, equalize the playback levels, and then instantaneously switch between them.
Believe me, I used to be a believer (along with an old friend, Kevin Voecks, the designer of Revel Loudspeakers), until I participated in really controlled comparisons of various amplifiers at the NRC in Ottawa. Kevin and I also used to believe there were significant audible differences in CD players (the analog preamp output sections may have errors in frequency that could be audible). After lots of tests that really bruised our egos, we changed our position.
Certainly power output into lower-impedances may introduce audible differences, but so long as the amplifiers (I'm only discussing solid-state designs), are not driven into clipping or near-clipping (distortion rises quickly as an amplfier approaches clipping), distortion is less than 0.5%, and the frequency response measures flat, then well-designed solid-state amplifiers that I've compared do not exhibit audible differences.
Axiom Resident Expert