Oculus was sold for $2B in 2014. Adjusting for S&P, I think it would have been sold for $4B today. MagicLeap has much more difficult and stronger tech so $10B seems reasonable. On back of the napkin, I would estimate ~500 people working for 4-5 years to build something like MagicLeap. So if a company wants to do this, they should be prepared to foot the bill of approximately $1B and wait for that many years.

The problem, however,is this. AR tech is simply not there yet. Glasses are becoming smaller but resolution sucks and we don't know if headaches will ever go away for more than an hour usage. I would estimate another 10 years of intense development before they are ready to compete with 6X cheaper 4K monitors in rendering quality. It's a long game and doesn't work without determined leader like Zuck willing to drain billions on it.