A nasty situation and, I guess, a problem at more than a few aerodromes around the traps.
The general philosophy involves -
(a) probably no consideration regarding body angles - consider that we have no great difficulty with different landing flap settings which entail different body angles. The reference to low vis considerations is a red herring, I suggest.
(b) flight test and AFM data is based on steady wind - one avoids gusty conditions like the plague due to the inability to obtain repeatable data.
(c) for a steady wind over a flat area, there is a ground layer (a bit like the boundary layer around an aeroplane in flight) which sees the freestream wind speed at a reasonable height decrease as the measurement is taken progressively nearer to the ground. Generally, this large scale wind shear is approximated by a standard equation routinely used in flight test work and usually presented in AFMs.
speed(1)/speed(2) = (height(1)/height(2))^(1/7)
This has been discussed at length in previous threads.
Obviously, this will be subject to all sorts of perturbations where there are obstructions. However, for most larger airports, the flat area analogy is a reasonable starting point.
It follows that this boundary shear will see the W/V reduce during the later stages of the approach and landing and that a modest margin for approaches into headwinds (where the headwind is expected to reduce) is appropriate but not for tailwinds (where the tailwind is expected to reduce).
This is all fairly predictable and repeatable. It is anticipated that the additive will bleed off as the aircraft approaches the runway and landing.
(d) for a gusty wind, on the other hand, the wind variation is unpredictable and random. Therefore there is a sound argument for carrying a measure of the mean gust through until the flare to guard against unpleasant surprises.
(e) the planned landing for heavies generally is factored by 1.67 on the flight test performance landing demonstrations in steady wind conditions. It follows that this factor ought not to be squandered lest one be bitten severely by the Fickle Finger of Fate. Thus there will be a maximum additive. Typically, this will be up to 20kt and this sort of value dates back to BCAR days as I recall.
(f) consider that, to a first approximation, variations in landing distance will be related to variations in speed squared. If, for instance, we are looking at a typical Vref of, say, 130kt, add 20kt with no wind losses to the ground speed on a bad day, and you might be looking at an increase in distance in the vicinity of around 35 percent - that 67 percent fat can disappear VERY rapidly and a routine landing becomes a major risk consideration.
(g) my view is that landing on a critical length runway in significantly gusting tailwinds is the stuff of emergency situations only and very sweaty hand business.