Please, I would implore any people seeking to measure the performance of anything "audio", to at least measure it in the context in which it is used, so that we don't have the mess that occured in the past.
I noticed that a very well-meaning poster above sought to use a steady sine-wave test tone and a load resistor in his testing. Is this not what got us into this mess the first time? Doubtless, negative-feedback-equipped power cords will result!
If someone is "hell-bent" on having some numbers to "lean on", then please try to procure them in a way that is meaningful to the way the product is operated when it is in its intended use. The idea(as I see it) is to find out something useful, and not to "make it easier" for the measurer.
Personally, I think that something may come of this, and I would love to see some new measuring protocols introduced that could address this common topic of why we don't seem to be able to measure the observations that we are obviously hearing.
You can't do this "in a vacuum". It needs to be done so that it measures the things you want to find out when the product is operating in the listening environment, playing real music. "In-situ", as it were. Test tones and "dummy loads" are out of the question. It is what happens in the actual reactive environment that we need to know, and test.