EDIT: Post based on the review here:
http://www.enjoythemusic.com/magazine/equipment/1202/bybee.htm
Obviously the reviewer was in direct contact with Bybee, and had some of the 'black magic' explained and/or contrived more thoroughly along the way.
I haven't found a single null test performed (and then explained away for lack of results :
) in almost an hour of leisurely searching. I think that speaks for itself. If we accept the assertions that:
1) 1/f noise is dominating the noisefloor in the targeted device
2) 1/f noise is frequency dependent
3) 1/f noise is not broadband, and has a well-defined spectrum
4) shot and Johnson noise are not being targeted by the device
then it stands to reason that it should be testable. If it's dominating the noisefloor, its absence should be rather conspicuous (all things being relative). Johnson noise in microphones strikes me as an appropriate correlation. If the 1/f noisefloor is significantly quieter in a comparable scenario, it wouldn't be perceptible anyhow- ergo the whole thing is moot. Frequency dependence means there's a nice variable there to tweak, and we're given where to look to prove claims (<2kHz). White noise isn't targeted, and since it's a theoretically consistent average throughout the spectrum it should either have no effect on testing or do it democratically.
Instead of this approach, all I've read so far are unsubstantiated claims and fanboy reviews. Not unexpected... such is the dark and depraved art of audiophilia.