Why dissonant music strikes the wrong chord in the brain

There's a slightly convincing example at http://sethares.engr.wisc.edu/mp3s/challoct.mp3 -- what you hear there is two successive sequences of (note A, note B, chord A+B) where B is an octave above A in the first case and about 1.07 octaves -- an octave and 0.8 of a semitone -- in the second. The instrument has a weird spectrum that makes the second A+B sound better than the first -- but it's weird enough that, to me at least, even the individual notes sound really odd.

A better example: http://eceserv0.ece.wisc.edu/%7Esethares/mp3s/tenfingersX.mp... is a piece in "10-tone equal temperament", in which notoriously almost all chords sound dreadful with normal instruments. http://sethares.engr.wisc.edu/mp3s/Ten_Fingers.mp3 is the exact same notes, but played on an instrument with a spectrum designed not to sound horrible in 10-tet.

Complex mathematical relations in harmonics induce more tension than simple ones. We've known this at least since Pythagoras. Schoenberg knew this perfectly well -- he published the bible for the theory of tonal harmony -- and never argued against it, as is claimed in this article. What he did claim was that our ears gradually grow used to the common dissonances which start to lose the tension potential and to be heard as "colors" [EDIT: hence one of his first atonal works, "Farben"]. This is what he calls the "emancipation of dissonance". Basically, we have to find the "edge" in more and more complex mathematical relations.

[EDIT: further, he thought this race for complexity was doomed and exhausted and we should let fully emancipated dissonance live a life of its own, not trying to build up and release tension; this opened a wealth of musical possibilities while alienating a good part of the audience]

Pure major chords sound more "pleasant" in a pepsi-challenge type of study. So does pure sugar in a taste study. That doesn't mean the perfect meal is a pile of sugar cubes.

There's also no discussion of cultures whose traditional music focuses on creating audible beating, like gamelan. There's a ton of audible beating, it's completely intentional and part of the sound, and even if you've grown up listening to music with a lower normal level of dissonance the sound can grow on you.

It's also worth mentioning that highly consonant music is pretty lifeless -- even in the world of western music, the drive and power of a piece comes in tension and release.

followed by:

Co-author Josh McDermott at New York University reported previously that harmonicity seems more important than beating for dissonance aversion in normal hearers. In the new paper he and his colleagues argue that the lack of sensitivity both to harmonicity and dissonance in amusics now adds to that case.

Am I thick? These paragraphs seem to directly contradict each other (1st says amusics are sensitive to dissonance, 2nd says they are not).

From reading this, you might think that this trick was invented in the mid-20th century.