I've just read this report from the DOE, and though it doesn't talk about windpower at all, I find it quite exciting for wind's prospects.
The conventional wisdom has been that the small size of the turbines (generally about 2 MW each) and the unreliability of both the wind and the turbines makes it improbable that the bulk of our power needs can be met with wind.
Meantime, the installed cost of windpower has been dropping, and is now at something like $1300/kilowatt of peak capacity, and coal-fired powerplants have been getting more expensive ($2200/kilowatt), and gas-fired powerplants have been getting more expensive to run (they remain cheap to build at $600/kilowatt). That doesn't explain everything, but check out this statistic from the DOE report:
From 2000 to 2007, the U.S. built an average of 293 MW/year of new coal-fired capacity. In that time, wind build rate went from essentially nothing to... about 4000 MW in 2007! Holy cow, that's an order of magnitude more build than coal!
Now I understand that, like the long Nuclear Pause, there has been something of a moratorium on new Coal for a (shorter) while. And, I'm told there are lots of coal-fired plants in planning right now. But just for scale, note that the EIA projects that the U.S. needs 6000 kW/year of new capacity for the next couple decades. Even assuming a 33% utilization rate, wind is within an order of magnitude of producing ALL of that new capacity, right now.
It's no longer a question of whether wind can ever dominate coal... it's a question of whether coal can come back! Look at figure 2 in the DOE report, and project a growth curve for windpower at 1300 MW/year in 2007 rising to 3200 MW/year in 2012. Why is my 2007 wind number small? Because you have to divide windpower by 3 to account for the wind not blowing much of the time.
Anyway, what you see is that wind will outpace coal again in 2008, but coal will win in 2009 and 2010. But after that, all this new wind capacity is going to meet most of the need for new capacity, reducing the need for new coal plants (and greatly increasing the need for long distance power lines at the same time).
And, by the way, there are about a dozen new nuclear plants in the works, perhaps half of which will come online in 2012 or thereabouts. They'll eat even more of the demand that would otherwise go to coal.
Here's a satisfying question to ponder: what year will U.S. coal production peak, not from lack of supply, but from lack of demand?