Gaston, Fechter, I agree that the cells will work at their best when they are warm. But we never have cold cell problems in Perth. Ever.
I've been at a track day at Wanneroo Raceway where the air temp was 41'C and the track temp was 52'C. Modern race bikes were overheating! So warm is good, but hot is not ideal.
e-beach, my preferred cells are 300 Wh/kg, 500 Wh/l, 90 C and $0.25/Wh.
Seriously though, I'm mostly using the 40C LiPo available from any reputable Chinese hobby supplier. These are built with cells flat up against each other, typically in muti-kWh packs. I have no problem with them getting warm, even quite warm, but if you were striving for the lightest pack possible for the application, you will be pulling high C's from the pack and getting it very hot.
I guess it would depend on where you can afford to hang a radiator off the vehicle. And also bear in mind don't want to cool cells down to a point where they will be less powerful, so perhaps one cooling plate every three cells would suffice? And lots of conductive bits of metal inside a battery is not going to make a crash particularly safe...
Edit - forgot to add, after a 13C discharge for 5 minutes (roughly 100% DOD) on a "well padded" pack we were seeing about 58'C. Not sure if this is too hot
or not, but the "well padded" is relevant here. The idea was that the foam matting would prevent vibration damage to the cells, but I suspect it was serving as a giant stubby holder, keeping it nice and warm. I also think the foam was way too thick for what we were trying to do.