The problem with a/b testing in a controlled group is forgetting what was heard previously. Take another example: if you a/b tested digital cameras or photo printers and you laid out all the samples in front of people they would easily be able to discern differences (provided they're not color blind, which is another factor - physiology). The greens are a little lighter, the reds are more vibrant, etc. However, if you handed out the photos one at a time, the results would be different. You may be able to recall differences with each succeding photo - but two, three or four down the line and you lose your frame of reference. Even if you observe each photo for an extended period of time, you still will forget.
Not the same with audio. You cannot listen to many power cables (or whatever) at the SAME time. You have to rely on your memory to discern differences since you are evaluating in succession. However, with the photo example above, just as professional photographers and graphic artists will be able to discern much more subtle color differences than the untrained eye, we can discern subtle sonic differences because of our listening experience or "training". But regardless of how good we discern sound (or color), relying on memeory is anything but accurate. And many times, we can only perceive the most subtle of differences only after living with a component for an extended period of time.
Nonsense, this a/b'ing IMO.
