I am not going to offer any argument or opinion on whether I think those papers that are published are real or just fear monger, mainly because I have no background as a scientist in the necessary fields. However, having worked with cellular base stations at my last job, I will note my own feelings towards it.
The important thing with end devices and basestations (whether that is for a cell phone or a WiFi access point) is coverage and signal strength. If you place a device on the fringe of a coverage area the base station and the end device will have to use basically the maximum signal strength to communicate. The same applies if the device is placed right next to the base station. This is equivalent to standing next to someone and screaming into their ear. Another is if there is so much nose in the area. A good metaphor is to compare it to a school cafeteria with everyone talking. At some point, it gets so loud that you have to start talking louder and louder just to be heard by the person standing near you.
Regardless of risk for RF exposure, with 2.4GHz devices it is good to scan the environment now and then and make sure that you aren’t operating on a frequency that is congested. It will reduce the overall radio power output, and also make communication much smoother, faster, and use less battery power.
With regards to “cooking”, which is something I am a little more worried about… Well, a microwave oven operates at about 800W or so, if I recall. My boss at my last job basically recommended NOT standing in front of our antennas when a base station was outputting at 50W (I also chose not to for the 10W configs). I did however routinely have a 100 mW desk unit on my desk. That didn’t worry me so much. I do know someone who was a broadcast engineer for KFRC/KMVQ/KYCY, and he told me that he would (crazy in my opinion) work around antennas that were transmitting (I forget the exact wattage he said). He’s kinda old and still alive, and this was in the '80s.
I’m not too concerned about devices in my apartment. I am only a little more mildly worried about constant RF noise that affects the health of an RF network; but not so much concerned about the power output being high enough to affect me.
That said… I don’t carry my cell phone in my pocket all day or talk on it a lot. I also don’t sleep next to my wifi router. If it puts off more than a few mW I try not to keep it “on” me for extended periods of time. In fact, keeping it on a belt clip is something I have considered instead of in a pocket (though my belt at the moment is kind of overloaded). As for cell phone base station antennas, the common set used by a cellular carrier is a sector antenna (they cover a varying degree, usually from about 60 degrees to 120 degrees). They are directional. So usually if you are behind it, you won’t get blasted. This is a good visualization (though it doesn’t have a “heat map” that would show signal strength at an optimal layout):
By contrast, and omnidirectional antenna (basically a pole – this is more commonly seen with transit and emergency setup, as well as TV and radio, but they tend to be up on taller towers for better coverage) would be a (lumpy) circle centered over the center of the graph on the horizontal graph.
Anyway, I realize that may be a bit scattered. But those are my feelings on the topic. (My irrational fear around this, though, are the full body scanners. But I think that is more of a “why are we doing this stupid bs?” thing. But, this is another topic not for here. )