Stall warning system a contributory factor?
To me as a lawyer with an interest in aviation safety, the preliminary report implies that the design of the stall warning system may have been a major contributory factor in this accident.
The PF was confronted with conflicting signals as well as degraded instruments. Nevertheless, at quite an early stage, his inputs commanded nose-down pitch. He appears successfully to have initiated a recovery. What he needed to do, was to develop this response further.
However, the stall warning system did not recognise this. It did not encourage the PF to develop his solution further. On the contrary, it sounded a stall warning. This is essentially because it could not reward the improving trend which the PF had initiated: it could only recognise the threshold above which the stall warning should be available (<60knots = stall warning unavailable; >60knots = stall warning available).
A correct interpretation of the stall warning would possibly have been difficult in any situation. Particularly in the difficult conditions which must have prevailed at the time, the PF may well have inferred that there was a causal connection between the stall warning and his most recent command: ie, nose-down. Naturally, he would have wished to reverse the cause. So he would command nose-up instead.
What is worse, the PF would now have learned by experience that commanding nose-down pitch would "cause" a stall warning. He would have been reluctant to try it again.
Therefore, my tentative view is that the stall warning at this stage in the attempted recovery gave precisely the wrong message to the PF. It may well have been an important step in the "critical path" of the accident.
I think we should not rush to blame the pilots: the term "pilot error" sometimes disguises the degree to which "human error" has been induced by automatic systems.