I find it hard to believe that the autothrottle response is so fast as to safely counter strong and gusty winds on very short final and flare. There is a huge difference between the published additive of only five knots to Vref with autothrottle engaged, and the up to 20 knots additive in the same met conditions but with autothrottle disengaged.
From reading accident reports where aircraft have over-run on landing in gusty winds and wet or slippery runway, in almost every case excess speed over Vref was a significant factor. It makes you wonder if the "excess" speed was in fact because all the additives (all the gust and half the steady HW component) came together and the planned-for gusts or lulls failed to materialise - resulting in a legal" excess speed" albeit a fatal one sometimes. All the gust and half headwind additives are used only if the landing is conducted without autothrottle of course.
Whatever happened to "Airmanship" when it comes to landing in adverse conditions? Seems that blind and rigid adherance to SOP is fine in theory but not to the stage where legal additives have the potential to result in unwanted float,fast touch downs and maybe over-runs.