Matt Hamblen at Computerworld does a nice job looking at the RCA/Audiovox Wi-Fi power harvester: This device will supposedly absorb Wi-Fi signal power from the air and convert it to stored battery power. This isn't an unreasonable notion, but it seems impossible.
The inverse-square law roughly says in this case that the signal would decrease in intensity inversely proportional to the square of the distance. Double the distance and signal strength drops by a fourth; triple it, and it drops ninefold; and so on. (It's got something in common with a soap bubble: a drop of liquid, when blown into an ever larger bubble, has less and less soap and water to maintain cohesion as it grows, becoming ever thinner.)
Thus at the distance at which most people find themselves in relation to access points, the amount of energy falling is incredibly minute. This is one of the miracles of Wi-Fi, that between using spread-spectrum and orthogonal frequency division multiplexing and a host of other tricks that a signal that is barely above the thermal noise floor can carry hundreds of megabits per second of data.
I like that Hamblen didn't dismiss the notion, but found reasonably skeptical people who explained the various reasons why it should be against the laws of physics to gather as much energy as the Airnergy device appears to be alleged to suck down.
The company made a splash by providing incomplete information and a seemingly specious number at CES--charging a BlackBerry in 90 minutes, although it may be that it took far longer to gather the energy into the battery by which the BlackBerry was trickle-charged.