"optical center" of the antenna probably is at least 5mm even with a phone pressed against your head, given the distance the ends are away. and things like clothing or purses add interposing distance too.
again though it's all a matter of how you design the test. if the source comes closer, the output is going to go up. using the same number but way closer than everyone else is actually using a lower number than everyone else.
again, I'm OOP and (while I'm not a doctor) my take is that the scale of cellphone ownership is far too big (at 4x+ these french emission limits, per the ANFR charts) for these to be reasonable decisions. cellphones at the customary frequency bands and emissions levels just are not that much of a health hazard, or else there would have been observable results bounded by the sample size (large). And really there are practical technical limits to how much power it's worth pounding out before you overwhelm the cell site or drain your battery in an hour. It's optimal for everyone if noise floors are low.
> "optical center" of the antenna probably is at least 5mm even with a phone pressed against your head, given the distance the ends are away. and things like clothing or purses add interposing distance too.
