One Outsider,
Whenever manipulation of a mechanism is required to achieve a certain outcome, the manipulation must be second nature so that the focus can be on achieving the desired outcome. If it is not, the focus will be on the manipulation and not on the outcome.
I believe you know this already.
Agreed, but where is the evidence that preoccupation with the "manipulation of the mechanism" played a role in these accidents? The initiator was a failure of monitoring, it was the autopilot that was doing the manipulation.
The incorrect stall warning recoveries were procedural errors. One was a very deliberate choice to hold a high pitch attitude exacerbating instead of resolving the condition. The other had a 7 second delay in application of TOGA power resultant from both a change of control and an erroneous autothrust mode that retarded power again after it had already been applied by the original PF.
There's nothing there to suggest that the way these stall warning recoveries were accomplished was the result of a "bucket full" situation with too much mental capacity being used for manual manipulation leading to a loss of focus on the desired outcome when the extra ball was tossed in.
ELAC