@fair
Thus, while testing on 2 ohm has its merit, it appears from the discussion that testing on non-purely-resistive loads is of more interest to people with practical experience in designing and repairing amps.
It is not in my review charter to help either camp in that. That aside, I did build an emulator of a 2-way speaker per stereophile. And used it for a bit of testing. It didn't reveal anything useful so I retired it.
I have used real headphones for headphone amps. That too was a useless exercise as the back EMF combined with the impedance of the amp produced completely erroneous and misleading THD measurements. Similar situation exists with testing amplifiers with speakers.
Even if something useful popped out, who is to say that speaker is representative of any other? It may be a corner case, an easy case, or a difficult case. Who knows.
I do vary the resistive load to test for load sensitivity and report on that (usually a problem with low-end class D amps although some high-end ones suffer the same).
Finally, keep in mind that any speaker or emulated load of one would have to be at very low power. There, distortion may not be material at all (swamped by noise). This is why JA at stereophile only uses is load emulation for frequency response test.
Yes, there is a $15,000+ load cube that provides a couple of reactive loads (NOT representing any real speaker). Audio critic had one but I think it blew up on them and they no longer used it. I would spend the money if I thought it would add value but it simply doesn't.
Making sure this point is understood: there is a very high bar for adding more tests to the suite that I run. Every test suggestion must come with strong justification which I have not seen in any of your posts. Folks on ASR routinely make such suggestions. You want a new test? Come back with real data that shows usefulness. "Would be nice" isn't going to work. A lot of things would be nice but not when it clutters existing measurements and take time and resources form testing other products.