Respect...
An impressive AAIB report which I am sure will reward more study.
For those unhappy that an extremely unlikely event should affect two independent systems within a few seconds of one another - there is a common factor over and above the shared history of those systems. The centre tank which is supposed to be empty, but is still in the loop.
The previous flight is significant because with ground temperatures in Beijing below zero, any ice from the outbound flight would not be scavenged.
The fuel uploaded in Beijing was warm enough to have melted this into droplets; the behaviour of the resulting droplets in terms of their size, accretion rate and so on, in the centre tank of an aircraft at rest in a below zero environment is problematic, but if it froze again before start-up, the water scavenge, and later the fuel scavenge would not get it. It would not start to melt and move from the 'empty' centre tank until the warmer environment of the descent.
So there is the credible possibility that at some time in the descent a melt surge presents ice, wateer and/or slush at the centre tank fuel scavenge points, and is then fed to the main tanks to be refrozen as it makes its way down to the boost pumps .....
If this happens during an unusually prompt arrival at late finals on a cold winters day, another hole lines up.
It would be particularly concerning if a procedural attitude or acceleration change could trigger such an episode which then prompts a temporary fuel restriction episode timed to coincide with a critical procedural need for more power.
Since the AAIB work is continuing, they must believe further recommendations are a possibility.