Based on what is currently known, this was a 'normal', SOP, accident.
The crew followed normal procedures and behaved as they had been trained and had practiced many aspects in everyday operations; except, apparently, for one oversight in not checking the thrust increase for an infrequently encountered situation, and a specifically exceptional circumstance.
If all preceding influences were to be known - acknowledged, then perhaps the industry could learn much more than cite the crew.
Consider:-
The tailwind approach was 'standard', and as might be expected resulted in an extended flare and longer touchdown than normal.
The crew followed the procedure for a RAAS alert, perhaps unexpected for the length of runway remaining. Was the alert distance optimised for shorter runways; what consideration was made for longer runways and, that with 'normal' tailwind landings alerts may be more frequent, - 'unwarranted' alerts. Were any of these assumptions published; more alerts were not expected, or accepting that there would be more alerts and 'GAs' (could this suggest ignoring the alert).
Was the context of the RAAS procedure fully explained / trained; as a Go Round, or as a Rejected Landing.
Does RAAS alert in the air or only on the ground. If airborne, selecting GA and then touching down is 'normal', the crew can depend on the highly reliable auto thrust system which is used every day.
If RAAS is ground use only then how might the crew identify the small window where the highly reliable auto thrust system cannot be depended on; is there a need for a 'different' procedure - RAAS SOP for Rejected Landing - not the GA SOP as suggested by the FCTM/FCOM text.
We have yet to see any pages for RAAS description and use.
Thus the crew in this normal, but unexpected operation followed the procedures, yet given human limitations apparently did not verify the thrust setting. The behaviour may have been normal; the AT was used routinely, presumable take-off thrust is checked, but with habit are the values actually read and compared with the expected value - a dependency on the very reliable AT system. Similarly for a GA, could the thrust increase and setting be as expected, apparently seen - 'wish think', and the mental workload compounded with over focus on SOPs. Selecting flaps, checking positive rate, mode annunciation, and then selecting gear up; all normal, trained for activities, as per the GA SOP.
How might this human weakness be identified in training. Did the instructors know about these human and technical factors.
Even if the exact conditions could be replicated, simulator training rarely represents the stress levels of real conditions. What chance of replicating the exact human behaviour; both pilots, at the same time. And where is the SOP for after selecting GA there is no thrust increase, ... Disconnect AT and move thrust levers (rarely if ever done).
This accident, even without a final report, is an opportunity to consider contributions and conjunctions, and thence what everyone might be review and as necessary change.