In reply to:

I'm just talking about the relatively smaller squiggles and humps in the middle of the response curves.

I think the missing point here is that the graphed data is captured by a machine, not by a human ear. While the machine can discern and display minute differences in frequency response, that does not mean that a person can actually perceive those differences.


Indeed Tom.
Those wee humps and valleys are often below the threshold of human hearing (about 1dB) But if a hump were sitting at say 0.98dB higher than the average, and a song was mixed 'hot' by 0.5dB at that same frequency, you now have a frequency point at which the human ear can detect a difference (theoretically).
Those small humps and valleys could mean more on a larger scale once all other parameters are put into place, recording quality being one major item of course.
I think the charts are more useful for looking at obvious humps and valleys, at least that is about all i glance for when i look at the frequency response graphs. I am still a bit surprised at the M80 graph. The M22 is pretty impressive. There were some NRC measures for some Thiels that looked downright amazing.

"Those who preach the myths of audio are ignorant of truth."