These numbers look impossible to me. 500 times more efficient would mean those other fuel cells have an efficiency of less than 0.2%. A quick google search reveals that mass-produced solar cells have an efficiency of 12-15% (with lab efficiency records of more than 40%).

"This solar cell can't be generating this much electricity, it can't be absorbing this much extra light," he recalled thinking.

I think so too ;)

12-15% is efficiency on the light hitting the suface, that are yellow sunlight, and not reflecting off

This is 500 times the total amount per sq in.

The 3D allows for more surface area, meaning more surfaces are available to receive light.
The cells are also able to receive multiple "colors" (light frequencies).

Check out the following article on spectrum efficiency:

http://www.newscientist.com/article.ns?id=dn3145

It seems to me like the best solar cells (Indium gallium nitride apparently) are already using most of the spectrum. But hey, I'm not an expert.

Keyword is absorption?

I like to equip Fire Robe and spam fire AOE on me to damage my enemies at the same time healing myself from absorption.

Those are not retail cells, notice the name of the article: "Solar cells aiming for full spectrum efficiency"

From the text: "It should allow solar cells to jump in efficiency from today's best of 30 per cent to 50 per cent or higher." This means that this is a little less than doubly efficient, 3D surfaces would improve that greatly.

Plus those cells use very delicate crystals that are slow growing and break easily.

I'm pretty sure (but not a scientist) that this kid basically is layering several layers of pretty standard cells that absorb different bands of the spectrum.

I'm sure 500 is an exageration.  And like I say on phys org all the time, the problem isn't the cell design it's the manufacturing and price.

To be fair by the age of 12 I had already built two robots (one of them could walk ). And I have often thought that solar panels are criminally innefficient because they don't use as much surface area as could potentially be used.

This kid is just like me... except he is smart.

Edit: I once built a solar powered RC car, but it was crap

I barely learned to wipe my ass in high school, and this kid is making privately and publically funded scientists look like fools.

Not until his invention is proven to work reliably and cheaply.

The kid is smart, no doubt about that... But these pop-sci articles often hype up things which don't deserve that much hype. I suspect that will be the case here, simply based on probabilities. In this case, the article doesn't even have any details on how his invention works.

How many times have we seen a world-saving invention hyped up in newspapers, only to never be followed up on?

Neat since he was 12 he probably wasn't smart enough to patent it,.

Only $25,000 Scholarship though?

Very true NJ5, but we never hear about those invetions and forget all about them :)