Cheers.
I thought that was the case, but wanted to make sure.
So - with that checked, your assertion about the discrepancy between the selected VS and actual VS during the descent phase of the sequence is correct (i.e. the FDs could well have been commanding a high nose-up pitch angle). However, during the descent phase with the aircraft stalled, the PF would have had a job keeping the nose consistently above 0 degrees, and chasing a positive FD command then would have been practically impossible.
What's interesting to me is what happened in the climb phase, and for me it's vital to keep these two phases distinct in terms of analysing FD behaviour and correllating it with the PF's actions. In both cases of the FDs' reappearance during the climb phase the discrepancy between the pitch angle, selected V/S and actual V/S should never have indicated an instruction to put the nose up to 15 degrees, but that's what happened.
Picking this apart is a complex endeavour!