The hypothetical question is, would a slightly lower faster approach, reducing the margin to the houses, have left the plane with enough additional energy to flare over the road and (presumably) come down softer than the actual 1400 fpm vertical speed?
Very difficult to say, due to the unpredictability of ground effect, wind gradient, drag characteristics of a 'dirty' airframe, engines that were producing significant thrust, etc. It's something that would require large gonads to experiment with in this situation... If it was *obvious* that a crash into houses/trees/petrol stations was going to take place, then what have you got to lose? Otherwise, a very difficult call. I wouldn't rely on a simulator to give a realistic answer in this scenario either.
It is well known that you can often escape an undershoot situation by using the technique described in the question above. What is slightly less well known is that the aircraft are generally "cleaned up" on recognition of a developing undershoot and the potential energy (height) converted into kinetic (speed), then bled off in the ground effect where the wing is more efficient. If you have gear and drag flap deployed on an airframe where configuration changes are slow and/or require a significant energy input, the act of speeding up to get into ground effect will probably bleed more energy than you'll save later and lead to an even shorter landing. It isn't called "drag flap" for nothing; in an overshoot situation, "pushing against the flaps" is a another well-known remedy.
A second question. What if the airplane would have "sounded the alarm" at the earliest possible time, which according to the trace was about 54 sec before touchdown for the right engine and 45 sec for the left, as the EEC entered "Control Loop 17"? An immediate flap retraction to 20 degrees at that early point in time might actually have given some real benefit of reduced drag, even considering the ~10 sec retraction time. But how much?
The were no warnings (aural or visual) that anything untoward was happening. This is just one of the factors that made it a very difficult scenario to deal with. Does going into "Control Loop 17" mean the engine has become unresponsive? I don't have that level of knowledge but I believe it has something to do with surge/stall recovery, not a total failure. From a flight crew perspective, as it happened to one engine first, a warning would be noted but I suspect no action taken, as the 777 is perfectly happy coming in single engine F30, as the WAT limits are high in that config. By the time a warning came through for the other one, you'd be at the same place and time as the crew were on the accident flight.
It is easy to take the precise figures for BA38 and work out all sorts of things the guys might or might not have done. Change those starting conditions by just a small amount and the "best" course of action may change radically... It was a very dynamic situation and there was not enough information available at that instant to make calculated decisions - just instinctive ones. After all, here we are,
eight months later, and no-one has definitively worked out what the optimum actions might have been...