Ahead of its time.
I really do think cheap generic hardware + streaming game service is the way things are going; it's interesting seeing the way it's developed -- first Phantom, which introduced the basic premise but never actually managed to produce a product, then OnLive, which put out a product but still couldn't sell it. Ouya's the next evolution; it's doing pretty well out the gate but the proof is in the pudding.
I've said it before but it bears repeating: in the coming years the latency problem is going to get mostly-tackled and we're eventually going to see consoles that are just dumb terminals that accept controller inputs, send them to a server, and then send back the audio/video/force output. This is desirable from the publishers' standpoint because it is the only way DRM can ever actually successfully work (because it is the only way that does not involve copying the program into the client's memory), and it's desirable from players' standpoint because it will mean cheaper hardware, up to and including handhelds that perform on par with consoles.
Ouya complicates things a bit because it's an open platform, which means fragmentation. The cost is a consistent user experience; the benefit is a competitive market of both hardware and software. I'm a Linux guy so I don't have to tell you which side of the divide I stand on, but monolithic It-Just-Works platforms certainly have their appeal.
What Ouya's got is Android, making for easy portability of existing games. Then again, OnLive ran Windows.