I can understand - not usually agree with but understand - how one might be ambivalent about listening tests even when they are carried out as rigorously as we can evision. What I don't understand or agree with is finding fault with comparison of electrical signals, such as null tests with osciloscopes. That takes the ear and brain out of the comparison process. Either the signals being fed to the transducers we listen to are identical or audibly identical by reason of level, or they are not. As for the usual counter about not taking the ear and brain out of the process, I call total BS. None of this is about the variability of hearing between individuals. It is about what the hardware and software are doing to the signal that will be heard by the individual listener.
When it was commented that if something was obviously audible it would be easily measureable, if the measure was electrical that WOULD be a slam dunk definitive. When the entire input creating the sound is being compared, there is no wiggle room about what can be measured and what can't be, or isn't being, measured. For that to be true something beside the electrical signal would have to be moving the transducer being listened to, and no one here is suggesting pixie dust.
I'm not sure if this was for the USB thread or this one, probably apropriate to both, but here it is.