mscar I can't answer your question but I hope someone can. It would be very interesting to see the data.
I would like to add my bit to the discussion regarding my own thoughts on crosswind technique, and possible performance penalties.
In my experience, it is best to use aileron in proportion to rudder imputs. If I have about half of full scale rudder imput, I would have the control wheel at approximately 45 degrees from horizontal.
I refine this too by looking at the ADI and simply keeping the wings level. If you look closely, you can actually see 1 or 2 degrees of roll as the upwind wing produces more lift. This is especially true for aircraft with sweepback wings. So if the crosswind is from the right, and the wings are rolling 1 degree to the left, I simply put in more right aileron and level the wings again.
With the correct aileron input as per above, the rotation can be done nice and slow (2.5 degrees per second) without any unexpected roll or yaw. The rudder and ailerons are already where they need to be. As a result, there is no urgency or need to rush the rotation and get the aircraft off the ground in a hurry.
Once airbourne and around 100 ft AGL, I simply slowly relax the rudder input over a 3 to 5 second period, and adjust the aileron input as required to keep the wings level. By the time the rudder input is back to zero, the aileron input is back to zero, and the aircraft is nicely tracking the runway centreline with the correct drift established.
Sometimes I get the feeling some pilots want to get the crosswind rotation over with as quickly as possible because they haven't got the correct cross-controls BEFORE the rotaion begins. They're just a tailscrape waiting to happen!
Regarding the performance penalty associated with cross-controls, especially aileron input and spoilers, I believe the performance penalties would be quite neglegable. For starters, flight control spoilers do not extend nearly as far as when they are used as ground spoilers even at full aileron deflection. Also the number of panels used as flight control spoilers is less than the actual number of panels present. Obviously they are all used when used as ground spoilers.
It has been argued that the raising of spoiler panels reduces the Vmcg speed. My answer to this is so what? Unless Vmcg is very close to V1 (which happens on very rare occasions) it makes no difference. Engine failure below V1, you close the thrust levers and remove the thrust asymmetry.
If you should have a V1 near Vmcg, you could reduce the aileron input for that take-off until well above V1 if you want to. But for the vast majority of take-offs, the benefits as mentioned above far outweigh the risks of negligable performance penalties.
Like I said, I love to see the performance data if there is any.