Wow that video is amazing if true, I'll be following its progress.

The actual power requirements for a flapping-winged aircraft are obviously difficult to work out because it depends on the efficiency of the craft (i.e. "sink rate"). But a way to take a simple view is to look at the efficiency of existing hang gliders and work from there as a best case scenario:

From

http://www.ornithopter.org/forum/showthread.php?t=429:

The best sink rate I could find for an existing hang glider was about 0.8 m/s at 8 m/s (17 mph). The glider was the Condor which is a large learning glider and travels very slowly.

This means the pilot would be effectively lifting their own weight 3 feet every second, which is about 1 kW. For human power you need about half that or less, otherwise it's too much for a non-athlete.

So in other words if you wanted to power a hang glider with a propeller, you'd need a 1 kW motor. That assumes the motor is 100% efficient, as well as the controller, wiring harness and propeller. Of course in real life motors are more likely going to be 90% efficient, controllers 95 perhaps and the propulsion is around 80% using a good propeller.

You'd multiply all those efficiencies together and get a value of like 0.90 * 0.95 * 0.80 = 0.68 . That means the whole system's efficiency is 68% . So to get 1 kW at the wing tips you'd need to feed it that 32% extra that gets wasted by the system which would be 1.32 kW input power for level flight (i.e. no climbing ability).

These figures are all rough of course and using the best case scenario of a well designed glider. But it's just to demonstrate a way of working through the numbers to help get a feel for the absolute minimum power requirements. For an ornithopter I feel the propulsion will be quite inefficient compared to a glider (I reckon only 50% at best) and the weight will be large. So if I had a gun to my head I'd choose at least a 10 kW motor.