An interesting topic and I have to admit switching from the 'limit control wheel input' camp to the 'use as much control wheel as you need' camp.
The reason for the switch was a recent B737 sim session in which exactly this problem was demonstrated and the wing lift with limited control wheel input on a 35kt crosswind was impressive and startling at the same time.
I fully respect Old Smokey's input regarding problems with vmcg and vmca with a low V1 but would I be correct to say that V1 / vmcg problems occur mostly on contaminated runways, short runways and low take off weights? Strong crosswinds and contaminated runways don't mix so this is an unlikely scenario, at low take off weights you can usually increase V1 = Vmcg without without compromising take off performance and for short fields there is no easy answer!
Statistical probability of engine failure at V1 is miniscule and on that basis I prefer to use the correct amount of control wheel input to prevent the rapid wing lift that will occur with a strong crosswind rather than fret about the possibility of an engine failure with a higher vmgc/vmca caused by spoiler deployment.
Happy to be corrected if I'm barking up the wrong tree.