And are you seriously suggesting that as a result of this accident, we should look at training crews in unanticipated double engines failures on Final Approach, or might it be better to stop the engines failing in the first place?
No. You might want to look at my post
#1736 to get confirmation on that point. This is a secondary issue, but not totally uninteresting, as many before me have shown

.
So you are suggesting that if the engines give a first indication of failure in an airliner, you want the crew to ignore all drills to do with trying to restore the engines, or even determine if the warning is/are false... and dive below the G/S, retracting flaps by (?) how much. And when would you do this? VMC? IMC? AWOPS? With Terrain under the approach path?
As BOAC has stated (and others before), there is an altitude (or point in the approach) above which reducing drag pays off anyway and I'm referring to that phase - I'm not sure this was the case for BA038. The suggestion would admittedly be bad in those cases where the time for the action (say 5 sec) delays the engine restoration drill AND the engines would actually have responded in the last 5 sec before hitting the ground. In all other cases it would do no additional harm. Either the airplane crashes anyway or the engines save it before, or - in the fortunate case - the airplane would increase its chance to make the airfield.
Take a clue from the AAIB - do they seem the slightest bit interested, at this stage, in the crew actions, or why the engines stopped
Finding the root cause is paramount. However, if it's not found with certainty, there might be reason to consider the handling of low level dual engines-outs. Time will tell.