[jillesvangurp]

True, it needs some infrastructure.

But on the other hand, this may be similarly complex/expensive as installing and maintaining a wind turbine on a tower that's 100m high is. These things tend to be pretty complicated and expensive as well. It's all about operational cost vs revenue in the end (and comparing that to other solutions). IMHO wind and solar are still making great leaps currently in cost effectiveness and efficiency. So you'd have to compete with/undercut that. Which is of course a challenge for just about any other form of energy generation currently (particularly anything nuclear of fossil fuel based).

The added advantage of this setup would be that you can land the things during a storm or for maintenance which potentially simplifies e.g. in the field repairs. Also, if they can be small enough that they can be mounted on a truck, deployment might be quicker or you might be able to re-deploy them to other areas. Currently building a new wind mill park is a large investment. Mass production of these things might enable more rapid roll out. So there are some upsides theoretically.

Safety might be a bigger issue. The cables would be a risk for planes and what happens if the cable snaps? Autonomous drones would be able to handle the latter probably (by crashing/landing in some safe spot). Also, lots of moving parts means those parts need more frequent maintenance/replacement. I imagine the stresses on the cable and mounts would be substantial (and proportional to the amount of power generated).

[jacquesm]

It is of an entirely different level of complexity.