You don't need hat and sunscreen for street lights or flashlights nor would you need it for cellular power levels, orders of magnitude difference in exposure.
Therefore, 99.9999038793% of the energy emitted is at wavelengths shorter than 1E-4m (0.1mm).
Or, across the entire continuous spectrum (0, 1e-4m], the total energy from the Sun is 0.0009612W, or 961uW.
You might complain that 961 uW is much more than 10nW, but again, you have to consider that 961 uW is across the entire spectrum (0,1E-4]. The Sun is less powerful at narrow spectrums because otherwise it'd drown out the cell phone tower (Or rather we pump power into the antenna to overcome the "noise" from the Sun).
I'd reproduce the numbers, but unfortunately my HP48 underflows at such narrow bandwidths.
And the "LED" at your face is, of course, much more powerful than the Sun since (as you pointed out) by the inverse square law, it has to pump out a lot of power to reach the tower.
Nor do I happen to think LEDs are harmless. Using LEDs to affect biological system is a very rich area of study.
Do I think non-ionizing radiation is harmful? I doubt it. But comparing a 1000W/m^2 @ 5400K black body radiator to a cell phone tower is dishonest.
Shorter wavelengths have higher photon energy. It's the short stuff that you have to worry about.
Is it consequential? Probably not, but the OP's original comparison of 1000 W of Sun vs. 100W from a tower and nW from cell phone is disingenuous.
The concern with 5G seems to revolve around the uses of higher frequency millimeter wave radiation compared to 3G/4G which has shown no repeatable damaging results at normal power levels.
If higher frequency = bad, which is actually true, comparing 5G the sun is not dishonest IMO. I am trying to give some perspective to show how it seems odd to worry about extremely low power cellular radiation while giving little thought to the extremely powerful nuclear radiator in the sky.
Dosage matters, cellular frequency will cook you given enough power, so will visible light. The power levels we are talking about do not generate enough heat to damage our tissue, so if they do harm it would need to be through some other unknown process. Should we keep looking for possible other processes, yes of course. However I would be much more concerned about the much more powerful visible range artificial radiators around us every day, like the monitor I am staring at right now emitting a 100 times the radiation of my cellphone right at my face all day long.
A simple power comparison is not a great measure of affect.
The problem with every argument about radiation from cell phones being dangerous is that for every proposed mechanism, the Sun is orders of magnitude more damaging.
The obvious worries with cell phones are repeated stress injuries, insomnia and disruption of personal relationships. There's just no plausible mechanism for the radiation to be damaging, though.