Sorry for the late response that you probably won't see. I could calculate some numbers, although getting an exact value is tricky because while with skin depth and antenna aperture you could relatively easily estimate the volume over which the RF power is dissipated into heat, I lack the background to calculate the rate at which heat will be conducted away and spread out. I'm sure that regulatory agencies have done this calculation.
You should really do some reading on the subject such as: http://www.durham.ca/departmen... (this is for Canada but from a quick google it seems informative, I'm sure there are similar US-specific references).
Anyway to give you an idea of a reasonable worst case, a typical phone or router will give off 1 watt maximum. Let's say for the sake of argument that you absorb all of that as a worst case (which you won't). Is that a lot? Not really. At about 4000 J/kg-degreesC for water it will take 160 seconds to heat one shot glass (40mL) of water by 1 degree C. During that time heat will be conducted away so you will likely see much less than a single degree temperature rise, which is safe.
If you want more detail there is ample scientific literature on the subject, but this back of the envelope calculation based on reasonable worse case conditions seems to show that the danger from RF emitters that are code compliant is negligible.