Ralph, I have read many dozens of your posts and I believe I have disagreed with you on very few occasions. But I don't agree with you here. If a system sounds better to me by matching dark sounding components with say, a pair of hot speakers then who cares about the measurements?
My point is that something dark isn't likely to be synergistic with something bright in an
ideal way. The reason something can be bright might be due to
unmasked distortion- higher ordered harmonics not masked by the 2nd and 3rd harmonics. We've all heard solid state amps with this problem! Since the
brightness isn't caused by a frequency response error, putting such an amp on a 'dark' speaker might result in a dark sounding system that is still harsh and annoying.
The measurements in this case
has to be the distortion spectra of the amplifier- if you only have 'THD' that's insufficient to make any kind of assessment, other than perhaps a really low THD figure might actually sound bad due to the only distortion it has being the higher orders. The ear uses the higher ordered harmonics to sense how loud a sound is, so its
keenly sensitive to their presence and if they've been messed with!
The reason you can get into trouble with synergies is the ear assigns tonality to all forms of distortion: the 'warmth' of tubes for example is the 2nd and 3rd harmonic having this assignment by the ear. Similarly, aliasing of digital is also distortion (in the analog world might be called 'inharmonic distortion') and similar to higher ordered harmonics, the ear perceives that as brightness. These assignments exist independently of tonal colorations brought on by frequency response error.
Mind you this is only one reason to avoid synergistic effects. If you're spending hard-earned cash, you can spend the same amount on equipment that doesn't have the problems (bright or 'dark') and so does not need to seek synergy. You wind up with something that is more musical IME.