Concluding the Great MP3 Bitrate Experiment

"I think it's clear from the basic summary statistics graph that only one audio sample here was discernably different than the rest – the 128kbps CBR".

Well, no, it isn't clear to me, at least not with the data in an un-sorted table. But visually, yes, it is clear: http://i.imgur.com/08s2S.png

well, I always yell at others for creating graphs which aren't anchored at 0, so here you go: http://i.imgur.com/7YrHi.png

Why didn't he randomize which order the samples appeared in?

tl;dr - 128kbps CBR audio has such bad quality nobody should use it, but all options of higher quality are indistinguishably different to the masses.

It's been a number of years since I compared, but I ended up doing rips at 320 VBR (or CBR -- I'm forgetting, at the moment) with the setting that allows bandwidth to fluctuate to whichever of the stereo channels momentarily has a greater, calculated need ("stereo" or "joint stereo").

Yes, the difference was noticeable, on a well-regarded iAudio player using decent can headphones. At 320, I was finally, really happy with the resulting rips. Though I'd still archive in a lossless format.