I did indeed read the article, and would say that it qualifies as a nicely run case study, nothing more. Could I draw conclusions from that work, and apply those conclusions to other cases with different listeners and equipment (i.e. headphones as opposed to loudspeakers)???
As an example, Axiom did their tests through their own loudspeakers. Each loudspeaker driver cone will break up into vibrational resonance frequencies depending on the design and materials of the driver. Did these driver resonances influence or mask the distortion introduced for the subjects to identify?
Did Axiom measure this resonance behaviour in their work to confirm it didn't influence the test? Doubt it.
Nothing Voodoo, just that conclusions from these type of simple comparisons are incomplete because the entire picture is not considered.
Again, that was a study to measure the level of distortion a set of listeners could hear from the signal.
The conditions were all the same for all the listeners.
If driver resonance would have increased the distortion, on top of the artificial one added to the signal, that would have decreased, not increased, the tolerance to distortion.
Unless of course, you are saying that those were somehow magic speakers, which were able to somehow remove distortion from the signal.
If yes, I think I am going to ask Axiom a sample, because we are onto something