cleeds,
"The tester’s bias - or even his personality traits - can poison the test."
I think you are taking this whole thing a little too serious.
Whole thing with double vs. single blind test is also a little exagerrated. I may be wrong, but it seems you extrapolate some, not even all, clinical medical research practice and apply it to something that does not require that particular level of stringency.
True, double blind testing will practically always be easier to defend than single blind testing, but in this case it is just for argument’s sake and not for any real world application.
As far as science 101 goes, thousands of mice die daily during tests and experiments that are not blinded at all, not to humans at least, and results are considered valid and used for whatever purpose intended. Ok, mice do not know much about it so you could say it is single blinded.
Subtle cues during some cable swapping experiment could happen, but it is expected that a person doing swapping would be an adult with at least some self-control. I guess you could call it bias, but it should be negligible. What is swapper going to do? Wink at the listener when installing a certain cable he wants to win? Clear her/his throat at that time?
Again, double blind testing may be ideal but it also may not be necessary.
In cable swapping test, single blinded with a reasonably behaving swapper may be all you need. Fully sighted may not be enough, if you are really trying to be scientific.