I do disagree when you say it is against all instincts to leave one thrust lever and retard the other one. It happened in at least 3 other accidents prior to this one and, MAYBE, because the thrust levers don't move in the airbus combined with the fact that the pilot was expecting to use just one reverser and to move just one thrust lever to the reverse detent, he was "mechanically" induced to not retard the other one, although we all know this is not the correct procedure. Just my .2 cents!
We're into semantics here
It would appear that all (?) of the problems have been with the 1 Rev U/S that generates this, seemingly unpredictable, pattern behaviour.
A lesson for the trick cyclists to look into, and presumably since it is so well defined, simple to address. The shame is that it appears maybe that was in hand, but not in time to maybe prevent this one?