You are making good points here, of course.
This business about biases & the need to control them always makes me smile.
What's happening in a blind A/B test when someone, like Amir, no longer perceives the difference he once heard sighted?
He now listens sighted & no longer hears a difference
So, now the bias that so strongly controlled what he heard before has suddenly been lifted & he's no longer under the control of the shiny, expensive bias gene - he now sees the 'truth'
Or is he now under the control of the 'no difference' bias gene which now determines his auditory perceptions for this device?
You know they say some people are easily hypnotised & some are not - perhaps there's a similar division in how people are prone to bias (or just how far they go to fool themselves that they are now hearing the 'truth' rather than substituting one bias for another?