One item I haven't seen in all these discussions is the flight directors. The report says that they disappeared along with the autopilot disconnection, and then reappeared and disappeared several times again. Crucially, the first time it reappeared it commanded 6000 fpm up, and the remaining times 1400 fpm up. Given the general FD overdependance talked about often on these forums, combined with the tunnel vision and stimulus-ignoring that happens during task saturation, this supplies a very big and easy puzzle to the "inexplicable" issue of why the PF kept pulling up.
I read this a while ago, and wondered why the FD would command this without being told to by the pilots. Just now I went back into the report, and that question is answered in section 1.6.9.2, which says that if they were not turned off manually, and reappear automatically when the computer data becomes valid again, they come back in HDG and VS modes. I'm sure this is common knowledge among Airbus pilots, which I am not. I can only assume that the VS value commanded is that which is what the airplane is doing the moment when they reappear.
So, if the plane was momentarily combing at 6000 fpm in a totally unsustainable zoom climb, it'll keep guiding the pilot to that climb regardless of the plane's ability to do that. And if the wide-eyed and overloaded pilot reduced his ability to observe (and interpret) all the pertinent display values, and reverted to trained habit (following the FD) then there's the recipe for the disaster that happened.
On learning this, I immediately thought that this is a terrible design wrt. human factors. The FDs, if off, should not come on again without their mode being explicitly set by the crew, according to what parameter the crew wants the FD to guide them to. Here, the "lock in the present VS" premise obviously failed the crew, who I'm sure were not aware of (much less in desire of) what they were following. Airbus pilots, has this been changed? What is the current behavior of the system in this case?