Yeah, Mark; one of the requirements for a properly controlled blind listening test is that the overall volume levels be matched to within 0.1dB. This isn't because a difference anywhere near that small is detectable as a loudness difference, but because in tests with certain material overall differences of just a few tenths of a dB were detectable, not as loudness per se, but as "clarity", "sound-stage", etc.
In the M60/M80 case, with the M80 being a couple dB more sensitive, not matching volumes as closely as possible(among other factors)might lead to describing as a qualitative advantage what was just a quantitative difference.
