Your method of frequency response has not been demonstrated at all to have value, again beyond the obvious cases of active speakers. Come back when you have proof of audible results, or objective proofs of audible compression and then we have something to look at. Otherwise it is a proposal for a measurement with shaky foundation.
My method is simply testing what the frequency response differences are as you vary the output. The exact same thing Sean discusses above in my topic. If you don't agree with it then you are welcome to discuss it further with him and Earl Geddes.
You may disagree. It doesn't make you right.
Now, let's imagine this scenario:
You test a speaker at 1v, yielding a mean SPL of about 76dB.
You test a speaker at a higher voltage, yielding 86dB SPL.
Yet, the difference in the result isn't a straight 10dB. It's different. Now, your bass output is lessened. What?...
Well, we
know - for a fact - that will alter the preference rating. How much? Well, no one has taken the time to do that testing. And it would depend on the severity of the non-linearity of the results.
Now, let's say two people are arguing about a speaker's performance. They both listened at the same volume. But one guy was 1 meter away while the other was 4m away. If we have proof that the response changes with output then we can work to make reasonable correlations about why the two had a different experience. That, at least to me, is the science of what we do as reviewers. That, to me, is the interesting stuff. Whether their experience be with passive or powered (limited) speakers. If they don't know the speaker output is limited or compressed then we are left to only look at other options or - worse - the subjectivist fellow of the two says "see, we don't hear the same" because they assume the SPL is linear... add 10dB, you get 10dB. But we know from the data it doesn't work that way.
Absolutely, most speakers don't perform terribly in this regard. But, on the flip side, we see HD increase dramatically in some cases. But it doesn't make the purpose of the testing moot. It just means that most speakers do well in this regard. And then we focus on other aspects.