Gosh, I'm a big fan of Nancarrow, honkey-tonk, parallel triads, and spanishisms notwithstanding; can easily listen to him for hours (I used to have a very short threshold for ligeti; couldn't take more than a few minutes before becoming very uncomfortable with it; I'm good now, though, I think).
I don't know all of Ligeti, by any means, but my threshold keeps getting shorter not longer.
As for parallel triads in Nancarrow, though they do have a connotation very inconsistent with certain avant principles, one could see them as more auratically neutral by suggesting that their voicings serve to stratify texture, nothing more (I think this is what Member Barrett was intimating as well). The same goes for octave doubled lines or lines doubled at the tritone, or any other such configurations. What makes triads particularly flexible in this regard, however, is the same thing that has always made triads more flexible than other sonorities: they divvy up the octave into roughly equal parts but with 3 unique intervals rather than (as in the augmented one) 3 equal intervals (major thirds being the same as diminished fourths in that case).
I sometimes like to say that the use of triads in traditional music is not motivated by any other factors either: you can move from one triad to any other one smoothly; You can't move from one <024> to every other one smoothly (CDE to FGA is going to require one or the other leap no matter how you arrange it). This is an entirely different take on the old saw that triadic music simply "sounds better" -- it also "makes better counterpoint". Or is that two ways of saying the same thing? I am fascinated by this line of inquiry.