There is no reason why excess heat couldn't be converted back into electricity but the effeciency of convertion will not be that high. I think theoretically the convertion would be around 50% efficient, but that's only theoretically mind you. Plus one has to realize that not 100% of the energy coming from the battery turns into heat to be converted in the first place. I have no idea how much of the energy in the battery is converted to heat but just guessing lets say 20%. Now for fun we'll assume the thermo chip is as efficient as it can be... well that leaves 10% of the batteries energy returning to the system. Now lets assume you get 4 hours of battery life normally and 10% of the energy is returned that gives you a total gain of 24 minutes. Now if this chip doesn't cost much adding 24 minutes is a no brainer, but i'd wager it wont be worth it. I'd also wager it returns less than 10% to the system.

AppleInsider:

"Ultimately, Brown said he sees Eneco's chips replacing heavy and bulky consumer electronics batteries altogether, but admits there's much work to be done to get the first demonstrative products build[sic]."

It seems that it's actually this Eneco guy that is making impossible claims. If I were in a meeting and he said this I would kick him out unless he clarified/corrected himself.

I can guarantee you Mr. Jobs is not talking to him at the moment.

By the way, AI also says that "The company claims its chip, which can operate at temperatures of up to 600 degrees Celsius, can convert heat energy into electricity at an efficiency of between 20 and 30 percent."

Sounds like thermotunnel cooling. They're claiming it works at 600C - exactly the minimum temperature for thermotunnel.

http://www.coolchips.gi/

Cool Chips use thermionic technology to deliver up to a projected 55% of the maximum (Carnot) theoretical efficiency for heat pumps.

No this is different than Peltier cooling. Peltier cooling uses the properties of different materials with differences in their conductivity to pull heat from one source to another (where they meet at two points) and can be controlled via the polarity of the current (that is which direction it pulls/pushes the heat).

In other words, Pelteir is moving heat from one area to another.

This sounds like it is creating electricity from heat, which requires cooling power. If the thing can create enough electricity to cool itself, it will indeed remove the heat without having to release it (like say, into your slowly roasting lap).

Apparently when radiation is butted up against a very cold wall (and very quickly loosing the battle) finds an easier release by converting back to electricity. I wonder if this gives us any new insights into he properties of light?

Not a chance. I'm as big a fan of "zero-point" or "free" energy systems as anybody, and I'm eagerly awaiting the day when somebody comes up with a proven way to get around the physical laws that have guided science to the present day. But this ain't it; a laptop requires far more energy than even a near-perfect conversion of the waste heat will provide.

Meanwhile, why don't we see photovoltaic cells on wireless keyboards, mice, and trackballs (and other peripherals)? In most cases, they could easily accumulate enough power in a battery while ambient light is high to keep running when ambient light is low.