Tornado energy has the same "harness" problem as lightning: There's a lot of energy there, but only for a short time. It takes a lot of expensive equipment to extract it and you don't get enough to pay it off.
Some of the L5 Society looked into windpower a few decades back, though, and concluded that you COULD do something with hurricaines.
By judicious modulation of ocean surface heating and/or evaporation you should be able to "Park" a hurricaine on (or just offset from) an island covered with wind turbines designed for the energy levels involved. Lightweight orbital structures redirecting a fraction of the sunlight from one part of the ocean to another could do it pretty easily - possibly as a directed side-effect of their own energy collection for the space-solar-power scenairos they were mainly interested in. (It only takes a little intervention to steer hurricaines: You're just tweaking the wind direction a bit.)
Think about a wind farm with a cross-section as wide as a small island and hundreds of feet high, in essentially permanent 100+ MPH winds (or at least half-year - and build another on the other side of the equator), and remember that the windspeed/energy curve is a cube function. A gross of nuke plants aren't in the ballpark.
Essentially what you're doing is turning a significant fraction of the ocean into a solar collector - with much lower material cost than paving it with equipment.
While shipping in the nearby area would be essentially impossible (except by submarine or by flying in over the storm and landing in the eye), the tethered hurricaine would also tend to "suck up" the vorticity of smaller ones, reducing (and controlling) the hazard from "wild" hurricaines on the rest of the ocean.
I don't know if anybody's done further studies on the idea now that we have supercomputing power good enough to model it properly. In the current political climate the environmentalists would freak out at the mere thought and some government would block the project. But it's interesting to imagine. B-)
