Solar calculations - part 1

This article is an interesting exercise in figuring out the scale of the solution. But I can simplify the math. He's basically saying the Nevada Solar One project cost $240M and puts out 64 MW of energy. That's $3.75 per watt. So he could have just skipped all his intermediate steps and said: $68 B (what Bush is asking for the Iraq war right now) buys you 18 GW. The U.S. needs ~300 GW for all our electricity needs. So $68 B buys 6% of our electricity. Amazing. Another way to think of it: with 300 million Americans, the cost to you is $227. That's it?? I'll see your $227 and happily pay my share of $3,750 to go 100% solar.

So why is this Economist article so pessimistic that solar won't amount to more than 1% of our energy needs in the next decade? There must be production limitations (i.e. we can only make so many panels per year currently). Because in that article they talk about costs of $1.40 per watt from cadmium telluride-based solar panels, which makes the cost per citizen to go 100% solar only $1,400.

Am I missing something?