Because if you used a sufficient number to light a room you would have to hire people to lift the weights continuously, and if you have to to that, it would be far more efficient if you hired them to generate electricity using a pedal powered generator, because that is still the most efficient way to transform human labor into electricity.
Come to think of it, pedalling for the time it takes to lift the weight and charge a supercap that will them provide electricity to the LED might still be better.
[Edit: Lets do the math: 100W effective for 30 seconds gives you 3000J electrical energy stored in the cap, drained over 3600 second provides an average of 830mW for an hour.]
Like others have said, this is in no way a replacement for a pedal-powered generator, nor is it a replacement for a standing lamp. It's for folks in developing countries who would gladly trade lifting a pound or two of weight for 30 minutes of dim light.
Be cynical about the math, sure, but if the math works, don't assume technology is useless because it doesn't have a place in a 25 year old's San Francisco apartment.
A pound of two? We are talking about 50 pounds, every 30 minutes, for very dim lighting.
It's not being cynical to say "It doesn't work". Cynical would be to dismiss it without even checking. We gave it a full consideration, ran the numbers, and realized: It doesn't, and Can not, work.
The numbers don't show that; that's your subjective judgment based on the numbers. But the judgment that matters is not yours, but that of the potential users. It's up to them to decide whether this is better for them than a kerosene lamp.
