Take away the cycling and you still have calendar losses, so I would say heat has its own loss mechanism. Granted, I haven't seen many studies of calendar losses without cycling (of course there are a few cycles to measure the battery), but
the one I have seen on chemistry similar to the Tesla shows calendar losses ACCELERATING once you pass a 15% drop in capacity.
Unfortunately, that is the regime the LEAF batteries in Phoenix are being operated in: calendar losses dominating over cycling losses. Does the LEAF have the same shape for its calendar losses? I don't know, but in absence of evidence to the contrary, I will suspect that it does. None of the anecdotal evidence posted on this forum to date leads me to believe otherwise.