Audio Research amps

[DrMorbius]

Seems both channels evenly matched at 8 ohms, which is my preferred setting.

 

[m. zillch]

Yes the blue and red almost perfectly overlap. That's what happens when people connect resistors to their amp rather than actual speakers which have a widely varying impedance per frequency, even if they (the speakers) may simplistically describe themselves as "8 ohm". They means that's their average/nominal impedance.
As I added in my edit it is the "simulated speaker load", the "gray" (looks black to me) curve, which I think would be audible with the right content.

 

[m. zillch]

Seems both channels evenly matched at 8 ohms, which is my preferred setting.

ALL of these various colors were made with the amp output set to 8 ohm. See my bold text of their statement, below:
""Fig.1 Audio Research I/50, 8 ohm output, frequency response at 2.83V into: simulated loudspeaker load (gray), 8 ohms (left channel blue, right red), 4 ohms (left cyan, right magenta), and 2 ohms (green) (2dB/vertical div.). "

 

[ski2xblack]

That blue is the other channel, but that's into a purely resistant load. That graph illustrates how higher output impedance results in linear distortion into a reactive load. Keep in mind that distortion can be broken down into linear distortion (freq response aberrations, which can be quite audible; that sort of nonlinearity can intentionally added via eq for that matter) and nonlinear distortion (harmonic distortion, which with the ARC amp should be negligible and inaudible).