http://res.cloudinary.com/engineering-com/image/upload/w_640,h_640,c_limit/PowerGraph_awx5de.jpg
http://www.engineering.com/ElectronicsDesign/ElectronicsDesignArticles/ArticleID/10749/Freevolt-Grabs-Energy-from-the-Air.aspx
Here's their white paper, where that graph (ie the first link, which isn't displaying as a picture, annoyingly - click on that link, or the 2nd link for the articles it's in) comes from:
http://www.getfreevolt.com/downloads/RF%20Energy%20Harvesting%20Whitepaper.pdf
However, that says the available power varies widely, which then seems to limit the use for the higher power items like smartwatches:
"Variations of power density were measured in a four-story
London office block. Measurements were taken across
different rooms and floors in the building as well as
numerous discrete locations in the offices.
Upper and lower WiFi frequencies had the highest average
power densities while the 3G (2100 MHz) and LTE (2600 MHz)
average power densities were the lowest. The highest peaks
recorded were in the sub-micro-Watt per square centimetre
range for both WiFi frequencies as well as the GSM/4G LTE
900 band, which were around 600-700 nW/cm2
...
The highest average power densities measured in outdoor
areas around London, were predictably in the GSM, 3G, and
4G bands. Peak power densities measured were in the
micro-Watts per square centimetre range for these cellular
bands, the highest being 6.7 uW/cm2 in the 3G band. In
comparison to office measurements, significantly lower
power densities have been measured for the WiFi bands
overall, with some peaks measured near tube stations.
...
the average RF density measured in an office or
external environment ranges from 20 to 35 nW/cm2"
which seems to put them at the low end of that RF range.