"Well, as you say, knowledge of the component under test does influence
perception. But so does being asked to make a conceptualized
discrimination

The first above has been demonstrated when humans are being tested in all
fields. The second has not been confirmed. When one of the hifi mags do
an "audition" and make reference to comparing this bit of gear to another
does this mean their perception has changed and their perceptions suspect?

"But how many "long-term, as relaxed as you like" comparisons have taken
place? Can you document, say, a dozen of them, along with the specific
conditions (length of time, directions to the test subjects)?"

Irrelevant, it is up to those advocating such a protocol to do so. Have
you
done subjective "tests" with your left hand embedded up to the wrist in
cheese doodles? If not how can you claim your "tests" legitimate?

In
science little time is given to considering that which experience in all
human testing has found consistent over and over again, knowing what is
being tested changes the outcomes. The alternative of the hifi mag
approach is almost completely useless in the extreme, except as a bit of
entertaining literary fluff.