I have owned plenty of tube amplifiers with high second harmonic distortion that sounded great, so a little different take on this than Paul.
If we have two amplifiers under consideration, and one has twice the THD as the other, does that matter? Would we choose one over the other because of its distortion spec?
I can recall years ago when Halcro trotted out their vanishingly low distortion power amplifier, setting the world back on its heels with triple and quadruple zero percent distortion levels—levels that challenged even the most advanced measurement equipment of the day.
And yet, the difference between an amplifier with 0.1% THD vs. another with 0.05% may not only be hard to hear, it’s entirely possible the higher distortion model might sound better. Or not.
The point of this is simple. When distortion and frequency response measurements exceed audible limits there’s an obvious result when we listen. When the measurements get below a certain threshold, those measurements mean less and less on their own.
To a design engineer, they might give clues as to the underlying circuits producing them, but used as a measure of sound quality, relative to others with very similar results, they are next to worthless.
Below a reasonable threshold, distortion measurements do not matter.