FlyingForFun,
Interesting comments & you sound like someone who has alot of experience of this. However, just a few more points. Forgive me for quoting you out of context or being too subjective but...
"The landing is made with wings level, both main wheels together, before the aircraft has a chance to start drifting".
"The landing is now made with the into-wind wheel first, followed by the other main wheel, followed by the nose wheel".
A bit of a contradiction in terms?
Before the a/c "has a chance to start drifting". That's rubbish.
Not even the best pilot can really everytime on a perfect roundout with more or less instantaneous heading alteration (rudder) and touchdown. Lots of xwinds variate and as someone said 'they reduce as you approach the surface'. Well generally, that is what is expected. Once well below the friction layer on approach though things could happen that are difficult to predict, especially at a point of pilot high workload.
Many xwinds variate through a range of degrees. Everyone has seen examples of this from the socks as they come in. Anything could happen. You may float, bounce, a gust may come along and tip your windward wing up, sudden windshear etc. I landed in approx 12 kts xwind the other day. The x component increased as I got to the surface! Generally unexpected.
Why rely on "chance"? Why not follow a preventative method that stops last minute drift and enhances stability? (Your correct in that you may realise there is very little drift during roundout hence so little windward aileron is required that your wings are almost level, depending on how much you are countering secondary effects). Why do you need to land with wings level if there is still drift?
Why not follow your second correct statement and land with windward wind down (hence windward main down as well). Gosh we even see the Saabs (commercial) doing it whilst we wait for RWY entry. Don't ever rely on chance unless there is a technique that removes it from the equation!