I may be a cranky old fart (in fact, I'm pretty sure I am), but this seems implausible (or impossible).
Yes, the rim will generate eddy currents when passing by a stationary magnet. This is a standard way to waste power in stationary bike trainers and even on dynamometers used on internal combustion engines. However.... they induce a DC current in the rim/disc, which will produce a DC magnetic field. As we all know, a DC magnetic field doesn't induce current in a coil of wire. It takes a change in magnetic flux to induce current in a coil.
Besides... the magnetic field induced by the eddy current merely acts to reduce the externally applied field. I'm at a loss to understand how anyone could extract any useful electric power from this arrangement.
Anyone know of any eyewitness accounts of this device? All I can find on the web is the press release from the inventor/promoter.
OTOH, if the rim had some steel segments built into the rim sidewall, such that 20mm was steel, the next 20mm was aluminum, the next 20mm was steel, etc, then you could put a magnet next to the rim. As the rim rotated, the field between the magnet and rim would vary as the steel and aluminum moved past. The field would increase when the steel was nearby, and decrease when the aluminum was near. If you put a coil around this field, the change in the field would induce current into the coil. This is a principle used in variable reluctance sensors, as commonly used to sense the cam or crankshaft rotation in an internal combustion engine.
I'd love to see a real advance in dynamo technology, but I won't be sending these folks my money any time soon.
regards,
Steve K.
edit: okay..... I've got working theory on how it works. It involves a magnetized disc that gets spun by the eddy currents dragging it around. A coil is positioned on the other side of the magnetized disc. As the disc spins, the changing magnetic poles along the perimeter of the disc induce current into the coil. I'm assuming that the two LEDs are simply wired in opposite polarity across the output of the coil.
In some regards, it's a lot like a bottle dynamo, except that magnetic fields couple the dynamo's axle to the rim instead of a roller.
The use of eddy currents produces losses in the rim though, so it does give up some efficiency relative to a hub dynamo. It's still probably more efficient than the average bottle dynamo, though.