Another important point is that it's not games he develops, it's game engines. Other people do most of the game logic and design. Working on a 3d display for those engines doesn't seem so far removed from that, and of potential value to future id/ZeniMax work.
He has made it pretty clear over the last few years that there really isn't a lot for him to do there, at least on interesting stuff. Writing rendering engines has become more incremental than revolutionary and I get the feeling that, for him, stuff like mega texture was about as interesting as writing a new zip program.
he used to be id first, armadillo airpspace second. now he's occulus first, id second.
it'll only preclude him from doing much programming.