... I simply do not believe you that the captain ...
When you study accidents and why things go wrong, you find that 'unbelievable' things happen regulary. A fraction of those become known as accidents. And a fraction of those are fatal.
Even very experienced professionals, who have no reason whatsoever to do something wrong, and certainly had no intent, have been involved in accidents that before hand where deemed 'unbelievable'. Part of how and why this happens fits under the label of cognitive dissonance. One clear and 'easy to show (low speed and 2D)' example that I have often used was a collision between a US Navy vessel and a Canadian Coast Guard vessel. These vessels were sailing on a separated parallel, safe, opposite 0/180 course. But in the end one vessel literally started 'chasing' the other that desparately tried to escape. In the end one the chaser hit the other port-aft with both vessels on almost the same course 135/135. Could you 'believe' the same thing happening on an airport apron?
An associated challenge (part of the overall challenge) for designers and testpilots is to decide which 'believable' and 'unbelievable' scenario's are tested against a possible design. I posted the example of a discussion between pilots and engineers about the access and location of CB panels. Believe has no place in such discussions. What you can use are scenario's and probabilities. And have these probabilities supported by as many facts as possible.
What we do not get in this case, with this low (CVR amongst others) content final report, is if the Capt did the 'unbelievable' thing of leaving his seat, how long that might have been, did he get back within these 9 seconds? or later, ... Answer to these questions would change quite a few posts and their implied conclusions. What this report clearly fails to deliver is as many known (CVR) facts as possible. And thereby fails to fully contribute to the improvement of safety.