Earlier quoted context omitted.
What gives you the impression that it won't be cheap to run? Presumably, if they've got the automation figured out, each unit will just manage itself, launching when wind conditions are favorable.
The underlying technology for the Makani kites and a standard turbine are the same. In the Makani case you have one kite with eight turbines, and like a standard wind turbine, all turbines need to be serviced. If you have to service 1 turbine, the other 7 stop operating. Hence, on a per turbine basis, I believe availability won't be around the 98-99% level (the industry standard), but lower. Additionally you have the…
Also, using smaller generators means that you've got a higher production volume, which can mean significant cost savings. You're also not dealing with giant bearings, gearboxes, etc. which are expensive to produce.
And if you're taking a kite offline, you might as well replace all the motor-generators at the same time. I assume they'll be designed to easily detach and replace. Then, back at the shop, you can check them out on your test bench, and see which ones can return to service in the spares stock.
I agree that the tether is a big question mark.