The very last study, on the very last page with the second paragraph
offers:
"For an effectively transmitted power of 0.25 W,
the maximum averaged SAR values in both cubic
and arbitrary-shaped volumes are, respectively, about
1.72 and 2.55 W kg-1 for 1g and
0.98 and 1.73 W kg-1 for 10 g of tissue."

The last study gives us more exposure data (the discussion of which
inevitably scatters in the rhetorical wind of debate). I can only
wonder if the reader can draw a conclusion from this quoted sentence
that can be expressed in temperature rise. There's enough data to do
this, only intelligence remains to perform.



that's pretty simple.. Assume that the tissue has the specific heat of
water. 1 Joule will raise the temperature of 1 gram of water about 1/4
degree C..

So, dump 2.55W/kg and you get about 0.0006 degree rise per second.
Hang on the phone for, say, 10 minutes (600 seconds) and you'll get a
temperature rise of a bit less than 1/2 degree C.

For comparison:
putting your head in sunlight results in an incident flux of about
1kW/square meter (peak). Assuming skin reflectivity of 0.36, the flux
being absorbed is about 640W/square meter. Let's assume that the energy
is absorbed in the first centimeter of your skin/bone, and that your
head is a circle about 10cm in radius (e.g. 314 square centimeters)..
That works out to about 20 watts total power being absorbed (compare to
the 0.25W RF in the example above). Again, let's say that the density
is 1g/cc, so the 20W is being dumped into 0.314 kg, or a SAR of 64 W/kg.

That's a rise of 0.015 degree/second, or 10 degrees in 10 minutes. In
reality, you won't see that much rise, because bloodflow carries some of
the heat away, and so does convection.