Consider that the evidence you are presenting does not show the actual influence of temperature on the virus itself. I would argue (anecdotally) that warmer climates mean people are less confined in general, therefore adding some element of "natural" social distancing.