I’ll take that thought a step further. ANY listening tests are suspect. Just try doing it yourself and you’ll see how fast you become overloaded and confused.
The only testing that has meaning would be done with scopes, spectrum analyzers and the like, and be free of any human perceptions. There are no audible anomalies that cannot be measured, whether it’s harmonics, clipping or any kind of distortion. And it should be measurable in the line signals before it ever reaches a speaker.
IMHO, even double blind listening tests are poor “science”, and their only purpose would be to prove or disprove that a measurable anomaly can be heard.
Medical science has been grappling with human perceptions for a long time, and the best research available in that area suggests that 40% of people will report changes in perceptions even in the absense of any intervention. If you apply that to listening tests, 40% of people will report differences when there are none.
Add to that the multitude of small physical changes that can affect our hearing, like body temperature, a slight tilt of the head or any change of body position, visual perceptions, and the list goes on.