Bear, based on the info available from BEA, which they look to have gotten from the heading trace from the FDR, there was no "rotation" but a gradual heading change of about 270 degrees (roughly) to the right. (In three or four minutes, to turn that many degrees I'll describe as "a gentle turn to the right" as an outcome (turn rate) but given the PIO early on in the upset, maybe that description is false).
Takata and a few others posted the picture a couple of threads back based on the BEA data/reconstruction.
As to PIO in roll:
It appears that the PF was able to eventually get a grip on that after his initial trouble in the roll channel. (Provisional verdict: the A330 is pretty stable in roll even at that AoA). I don't think I am the first to consider that his initial difficulty in controlling his roll once in "Alt 2, here, you have it," may have contributed to a scan breakdown regarding pitch.
We can't read minds, no less read minds from over two years ago, but JD-EE has raised (in the other thread, I think) the suggestion that in the back of his mind, he was worried about descending down into the goo he was flying in. We'll never know.
Insofar as
'Approach Turn Recovery' should that not also include a Rudder correction for the crab that remains? ~lyman~
We taught a balanced flight recovery, meaning center the ball as you fly. The approach turn stall was entered in balanced flight. We had another maneuver we taught, a skidded turn stall, that was a similar maneuver with the ball out, which tended to depart from controlled flight dramatically. (Not uncommon to roll over during the course of it). It was, of course, practiced at altitude.
Rudder was necessary to keep ball centered, particularly as in the prop plane adding power typically yaws the aircraft and must be countered with Rudder. (Heh, one of the first Approach Turn Stalls I ever did in a T-28B as a student, I torque rolled due to not stepping with
elan upon the rudder as power was added. Got me an earful, and rightly so).
Did so many of them, I guess that for me it sort of goes without saying.
(Again, this practice maneuver was practiced in VFR conditions).
In Re Garage Years point on thin air:
Both control authority and dampening would be attenuated in thin air, hence the potential to over control.