I am intrigued, and somewhat surprised, by the emphasis by some on the Flight Manual information error, in which a stated 3-4 minutes difference in fuel starvation times actually became around 30 seconds.
Surely to reach that point and expect to make use of it is the last desperate hope. I find it hard to believe that anyone views that error as significant in any real flight safety sense - given the quite clear 'land within ten minutes' (after low fuel warning) instruction, also in the FM. I really can't believe that was in any way in the pilot's mind.
I don't really understand the point you are making. I am not suggesting for a moment that it was in the pilot's mind at any time leading up to the first flame out that he could rely on 4 minutes extra time. But once we have reached the point of fuel starvation and first flame out, whatever the reason, it makes no sense to design the relative sizes of the supply tanks so that the second engine is starved after as little as 30 seconds. They should be designed such that the second engine has at least a few minutes fuel under all circumstances, which also allows critical items like radio-alt and search light to continue to be powered. That would at least give the pilot some chance over a built up area in the dark, rather than everything shutting down almost instantaneously.
There is another point that has not been commented on. In the Feb 14th 2014 special bulletin the AAIB stated that both prime pumps were ON. They have now back tracked on that and say on P46:
"There was significant displacement of the cockpit overhead switch panel
which appeared to have been driven downwards on to the upper surface of
the instrument panel coaming. Photographs taken by the first responders
appeared to show at least one of the prime pump switches in the off position.
When examined by the AAIB after the accident site had been stabilised, both
the No 1 and No 2 prime pump switches were found to be in the on position."
Now to me that calls into question the reliability of the assessment of the position of all switches, particularly in the overhead cockpit that has sustained serious damage and potentially been interfered with by untrained first arrivals. It also potentially blows out of the water the gathering consensus that PRIME and TRANSFER switches were mixed up by the pilot.
Take a completely different scenario where the pilot realised at the very last second, perhaps at the moment the first or even second engine flamed out, that he had the transfer switches OFF and switched the transfer switches back to 'ON' but too late to recover the situation and too late to stop the engines flaming out. And suppose when the scene is examined they were indeed found to be ON. Would we ever have known that they were OFF right until the very last moment? Relying on switch positions after a 100G event seems a leap of faith.