... The test simply made it difficult to identify the differences because listeners are used to hearing samples in sequential order and didn't conform to the standard pattern.
And it confirms my suspicions. When normal visual or audio cues are absent, such as the subconscious SELECTION BIAS that occurs when you can physically view the switch from the Kemper profile to the Reference Amp (e.g. YouTube video with annotations), and all you have is a randomized audio selection with no obvious visual or audio cues...even the experts with "golden ears" are fooled.
Now, when all is said and done, the mere fact that the Kemper can accurately clone the tone of the reference amp with authenticity and fidelity, would be rather pointless (for me), if it didn't also capture the just as important (perhaps more important) interactive dynamics between the guitarist and the amp. Here, I am speaking of things such as response to pick attack, note articulation, feedback, sustain and natural note decay, response to guitar volume roll-off, harmonics, etc., etc.
IMHO, it is precisely because the KPA succeeds on these subtle but oh so critical factors, that it separates itself from, and rises above the landscape of alternative digital modelers.