The amount of resources needed has been greatly exaggerated in the media and in popular science fiction. In reality, a 1.0 AU bubble-type sphere would require around the mass of the second largest asteroid, whereas cutting the size down to the orbital distance of Mercury would reduce the cost to a tenth of that. (Really, you'd want to bring the collection surfaces as close to the sun as possible.)

And remember, it's a power collector, not a habitat -- you could beam the energy out to habitats wherever you want throughout more distant space. Smaller spinning containers are much more efficient for producing habitable area, thanks to the square-cube law, and much easier for controlling things like weather.

More likely than putting that all towards landmass production, you'd put the power to work on some scientific project or another. I am a fan of the idea of using it to power vast fleets of computers, and using that to crack hard problems like discovering a cure for cancer.