As others have mentioned, the air increases in speed because of the Bernoulli principle: the decreased cross-section increases the dynamic pressure. However, most of the power will be lost in friction, during the "channeling" part -- and the increase in speed will not make up for that loss.
The friction is relatively high because the flow is turbulent, not laminar. The friction factor can be calculated with the Colebrook equation, implicitly; if they did the calculation they would realize that the distance between the intake and the turbine is a mistake (especially the height difference, which doesn't make sense for atmospheric flow).
It would be better to use natural land features that create the kind of flow you need, such as mountains. Look at the figures here for inspiration:
The friction is relatively high because the flow is turbulent, not laminar. The friction factor can be calculated with the Colebrook equation, implicitly; if they did the calculation they would realize that the distance between the intake and the turbine is a mistake (especially the height difference, which doesn't make sense for atmospheric flow).
It would be better to use natural land features that create the kind of flow you need, such as mountains. Look at the figures here for inspiration:
http://journals.ametsoc.org/doi/pdf/10.1175/1520-0469%281973...