I've used the 'crab' technique in everything I've ever flown, from Cessna 150, Chipmunk and PA-28, for example, to Buccaneer, F-4, Vulcan and VC10.
One very important point is that the approach speed must not be excessive if the aircraft is to behave as the designers intended in the flare. Use POH values, do not add an extra 5 knots for the wife and kids and all will be fine.
Otherwise, as you align the aircraft with the runway centreline during the flare, it will float and drift off downwind, which is extremely difficult to correct neatly.
There is NO 'kicking' in the flare - except afterwards by the instructor to the student if he used excessive rudder inputs! One flies with the correct drift applied to track the runway centreline, then yaws the aircraft to align it with the runway at the same rate that the flare is initiated, maintaining wings level as necessary with aileron. The VC10 had pretty heavy controls (requiring both hands on the yoke and power settings called for the flight engineer to set) and a powerful rudder system, it also had a significantly swept wing which would cause excessive roll if yawed too abruptly. Although the aircraft had a 28 kt dry crosswind limit, I once landed on the limit at RAF Mount Pleasant and found it quite demanding, but straightforward enough. I asked for the wind readout after landing and ATC advised that it had been well over 35 kt directly across the RW.....
I've had a few demonstrations of the 'wing down' technique in light aeroplanes and found it thoroughly unnatural and unnecessary. Of course those who maintain excessive IAS during the approach and then fly one wheel onto the ground, waiting for the speed to decrease before levelling the wings, can do so if they feel so inclined - but why? A wings-level balanced approach tracking the centreline, followed by a smoothly controlled 'de-crab' and flare works every time if the correct approach speed is used.