I'm not sure the 'negative elevator effect' was ever a practical issue in flight (as opposed to on rotation) - the response was entirely normal and I reckon that the increase in lift is near enough instantaneous. Aircraft with tailplanes would have a similar theoretical effect.
I don't recall hard landings being an issue on windy days - quite the opposite.
If there was an influence due to wind I would say it's more likely to be that in the gusts one may be tempted to 'tweak' the attitude: Putting the nose down by half a degree at 50' would have very disappointing consequences....
From memory, we had 4 speeds available:
Vref 'Normal' final approach speed (actually not used that much)
Vref+5 Engine out speed
Vref+7 Noise reducing approach speed (used as often as possible)
Vref+10 If the headwind component was over 25(?)kts
Most were at +7.
Vref was least nice - you had a higher attitude to start with, and needed more flare, which meant tail clearance was tight. It also meant that if you picked up a high RoD at 50' or so, it was VERY difficult to catch. One *could* add a bit of thrust at 30' or not close the throttles at 15', but this was not without drawbacks.
Vref+5 was better - the reason it was used was to give better g/a performance on 3 engines (there are obviously a lot of square laws at play here, because it made a significant difference).
Vref+7 was used off pretty much every ILS approach to a decent length of runway, where we would carry out a noise-reducing approach. This is probably explained elsewhere in the thread. These days it would be called an unstable approach! It gave you more lift margin into the flare and also more room to make pitch inputs (i.e. space for another half- to one- degree of flare).
Vref+10 was for windy days. I liked it! If you arrived at 40' in the right place, you basically just held the attitude and the ground effect did the rest. It did still *feel* like you had flared, as the ground effect would push the nose down and so back stick was still required to hold the attitude.
I hope that has answered CliveL and Megan's questions somewhat?