http://en.wikipedia.org/wiki/Bernoulli's_principle (Search for [18] in the Wikipedia article. Note that if z and p decrease from the air being lower and in a smaller pipe, then v increases.)
The reason to make it go faster is that you can't spin a turbine to generate electricity if the wind is 2mph, unless the turbine is really, really big.
It's an interesting optimization, but the broad design of the tophat might face challenges from adverse weather. During hurricanes or say something like 100mph wind, the upper flower has to recede to save itself. This is not the case with standard wind turbines that just rotate only faster. I am sure they must have thought through all these situations, simply considering only 600% improvement over standard installation doesn't seem like a good way to compare.
Standard turbines are not left to rotate in high winds -- if they would be, they would destroy themselves. There's a link to a video of this happening in this here: http://www.youtube.com/watch?v=_e9uRSVun30
Instead, they are feathered and locked in place. That is, the angle of attack of the blades is reduced to zero so that wind hitting them won't make them turn, and then there is a huge brake in the system that keeps them in place.
There are severe economic issues with paying for a giant funnel using the wind it generates.
No one has discussed if you basically turn the funnels into cell tower antennas then you piggy back onto the existing cell phone ecosystem.
Or piggy back onto land mobile radio trunking system ecosystem, or legacy broadcast media ecosystem.
You're building a giant horn antenna. Why not make it conductive and collect some "free" cash?