Sounds like a great technology to me, but with some caveats. The collectors would be very high efficiency (no atmosphere to deal with), but the capital expense must be enormous. Especially since they'd need to be deployed higher than most satellites (I would guess near or higher than geosynchronous, which is ~38000km), otherwise they'd suffer the same nighttime occlusion that plagues earthbound solar collectors.
Oh, and while collection would enjoy high efficiency, transmission would still have to deal with atmospheric effects. Also, transmitting megawatts through the atmosphere is bound to have side effects (I would think the heat bloom would be substantial), but maybe they've got a solution for that.
Edit: Meant to quote this...
Unlike ground-based solar arrays, space satellites could generate power 24 hours a day, unaffected by cloudy weather or Earth's day-night cycle. The capacity factor for a ground-based solar is typically less than 25 percent. In contrast, the capacity factor for a power-generating satellite is expected to be 97 percent, Marshall said.That's the issue that virtually all solar power promoters elide. They focus on max power production, and ignore the average, not realizing that the nature of solar power and Earth's curvature require you to provision two to three times as much solar generation as you would coal- or gas-based.
When you consider the number of collectors you have to build, as well as the number of power lines, solar looks a lot less environmentally attractive. The space-based solution goes a long way to neutralizing that issue.
(And yes, I realize the article says solar's capacity factor is less than 25 percent, suggesting you'd have to build four to five times as much solar capacity. That would be true if power consumption were uniformly distributed over the 24 hours in a day. But power consumption falls significantly at night, meaning nighttime areas wouldn't be drawing as much, requiring less power capacity.)
