Difference in rollback times
The AAIB report records, but does not comment on, that the fuel flows to the left engine were generally lower than those to the right engine, whilst the EPR of the right engine reduced first and to a value slightly higher than that of the left engine.
If, as is hinted, the engines (or rather some of their pumps) were separately sucking fuel from the pipework downstream of blockages at similar points in the two sets of pipework, it seems reasonable that the less thirsty engine could run for a few seconds longer. The extra pipework leading to the APU could possibly give the left side system a little more volume to suck from - but this would depend on the position of the constriction.
However, whilst possibly explaining the time difference between the rollbacks, I might expect that with similar restrictions the less thirsty engine might develop the larger, rather than smaller, EPR.
On a different theme, a couple of questions that physical chemists may be able to answer with a definite NO: does the presence of dissolved water alter the increase the temperature at which some waxing compoents solidify; and do the variations in composition of the Chinese fuel from normal Jet A1 modify the response to the previous question?
And finally: would it not be possible to provide the QAR with enough battery or even capacitor backup energy to enable it to write the volatile memory to non-volatile memory on failure of external power?