You may want to read the measurements and Koch's response before you lump him in with the snake oil boys. He does have a little experience in digital audio. What he said, at least to me, makes theoretical sense. If you optimize your algorithm for the reproduction of acoustic signals that are a complex mixture of tones and then measure with a single pure signal you may not be measuring it properly. However, I do not see how that explains the specific measurement anomalies that JA found in this case. That could also be because I do not have Koch's expertise in the field or it could be because he has a vested interest in the product, or both. That is why I found his response interesting.