Relax, I'm on your side  If you can't tell, it's either because I'm trying to take a relatively balanced view, or because I'm not clear enough - and it's difficult to tell the difference between those two possibilities...

Re Leventhal, I think you're trying to pick an unnecessary semantic fight with me here. To be entirely precise, yes, his argument began from the premise that the probability of detection is small, but it's surely not unreasonable to define a "small difference in quality" as a difference for which the probability of detection is small. Leventhal makes essentially the same assumption in his own preface to the Stereophile discussion:

 I do know that many listening tests using the ABX comparator... are conducted and analyzed in such a way that subtle differences [my emphasis] actually heard by the listener will likely go unidentified by the experimenter when the data is analyzed. 

For those who haven't read the discussion, Leventhal's point is most emphatically not that DBT is invalid, simply that many of the extant statistical analyses of the results were flawed. He wanted more DBT results in order to be able to draw any kind of definitive conclusions - not to junk the whole idea of DBT as the Stereophile editorial people seemed to think... And on the subject of Stereophile's "interpretation" of Leventhal's work, I'm inclined to think that their abuse of his results arises from the writer's incapacity to understand them rather than - as you suggest - a cynical manipulation for their own purposes. Either way though it's pretty irresponsible for them to publish nonsense like that on a subject which is obviously close to many people's hearts.

Perhaps for the sake of easier discussion we could call a difference which is very readily detectable a "large quality difference", and a difference which is detectable only with difficulty a "small quality difference". That would still leave most of the subject matter under debate under the category "no quality difference "  And if you read my post again you'll see that I made a similar comment regarding the claimed huge differences found under subjective testing conditions.

But I see the subjectivists have resorted to avoiding the debate through trying to ridicule it, so maybe it's heading towards time to wind the thread up