I said listening, not measuring, not objective, not blind.
What Hi-F describes their listening process here.
I am not looking for anyone here to endorse that process.
I am asking whether these amp characteristics can have any objective basis, or are they all the result of fertile imaginations.
Amplifiers of today, and since about the mid 1960s vary little. If you terminate the input and have no signal, you will have some variation in mains hum and broadband noise.
There is a slight variation in overload which should never be present. It will be similar in character across groups of amplifier topologies, including the negative feedback topology.
Theoretically you could build a pure resistance attenuating network between the amp and the speaker that would present about the same original speaker load while diverting power from the speaker, then listen to the amp near overload through the speaker. That would be like listening to the knee of
@amirm's distortion verses power graph. The idea would be to listen in steady state to what would only be found in peaks where there is not enough power for the speaker. But today in mastering the peak to average is very controlled because you never want to hit full scale on your A to D converter.
They would be more credible if each reviewer had an ongoing log of live music they are listening to with details. They could list their reference recordings often discussed here. The other problem is the deterioration of hearing in the reviewers with age. Their "reference equipment" except the ATC studio monitors show a bias for cost and snake oil. Using the same listening room is good. Adding blind A-B testing would be a good idea.
From the early HP distortion meters and their famous oscillator, work by Brüel & Kjær, and Tektronix - birthing Audio Precision, test equipment was the realm of engineers.
These review companies are a business, not science or engineering.
TL : DR Bollocks, complete rubbish for electronics.
