Not sure if this was posted before, but here are some numbers from the testing:
http://wikidevi.com/wiki/ASUS_RT-AC66U
from there you can go to this page:
https://fcc.io/MSQ/-RTAC66U
at which point you can look up this link:
(click on "Details" of the provided links on that page)
https://apps.fcc.gov/oetcf/eas/repo...me=Y&application_id=365917&fcc_id=MSQ-RTAC66U
There you will find a report with test results for this router. Here's that report:
(once you are in "Details" page following instructions from previous step, click on "RF exposure" test results)
https://apps.fcc.gov/eas/GetApplicationAttachment.html?id=1709263
Another link that should get you there:
https://apps.fcc.gov/oetcf/eas/repo...e=N&application_id=929239&fcc_id='MSQ-RTAC66U'
Looks like they measured the antenna gain to be ~2dBi (combined). Looks a bit low, but possible.
The reported maximum power on 802.11n (2.4GHz) is 640mW (sent into output antenna) with power density result of 0.2mW/cm^2 (at 20cm distance from the router antenna). Not sure if they somehow took into account antenna efficiency as i couldn't find any references to antenna measurements. Nevertheless, according to the report, passing value must be <1mW/cm^2.
For 802.11n on 5GHz band has shown max power of 800mW, which gave power density about 0.25mW/cm^2.
For 802.11ac on 5GHz band, the maximum power was 760mW with power density of 0.24mW/cm^2.
If we assume that all three standards and 802.11g (338mW) are blasting at maximum capacity simultaneously (or asus decided to increase power on one of the standards), we get (approximating with linear power addition) 2540mW of power blasting into output antenna (i.e. ~2.5Watts). using the same equation that was presented in the report, this gives about 0.8mW/cm^2 power density (antenna gain doesn't seem to change much for different measurements and stays about 2dBi). It is still lower than 1mW/cm^2 required to pass the test (at 20cm distance from antenna).
It would take about 3.17W of power to get to the test limit (assuming antenna gain and PA efficiency don't change - unlikely in practice), i.e. almost 25% higher than the max theoretical power during the FCC compliance test. It is true that no test should get to the limit, but looks like the router did have some margin which might have been partially utilized for initial production units (to gain good reviews?). It is not known whether they used worst-case units to conduct the test, which is why I would personally like to see some margin remaining. The retest with latest FW (which reduces power allegedly) still shows asus ahead of the pack on long reach up/down links. I may have incorrect info, but even from SNB retest article and some user comments it appeared that the power reduction was ~10-15%? -- need better source to quantify it. If asus was scared of possible scrutiny, the latest FW release must have brought the power back down to what was used in FCC compliance test or a maybe just slightly above it. If so, it looks like the only claim netgear has against asus is that production units are not the same as FCC compliance test devices, even though they all should still be within or close to the limit.
Summary:
Overall the numbers indicate that initial compliance passed with margin and even if asus did bump up the power on production units, it would've probably still pass the FCC compliance for most devices, but with a smaller margin (if any). Doesn't look like netgear will have an easy time winning this case.