This is an interesting point. The fact that they need 10k people to work on this makes them more like a government putting on a make work project than a company doing R&D. They'd be better off funding 1000 10-person startups working in VR, or just buying or copying one when / if it gets traction.
Related, I can see some appeal of getting to work at a small group doing development in a new area. It's got to be much harder to attract talent who know they are joining as a cog in a 10,000 person centrally planned effort