[probablybanned]

After thinking it over a bit, I'm going to take the other side of the bet.

We agree that power consumption is a limiting factor no matter what. In my imagined system, only something on the order of 1W would be delivered to the glasses. Delivering 1W over 10cm with large-ish coil area at >50% efficiency using NFMR isn't really a breakthrough, but it's only a way of eliminating the LiPo bulk and nothing more.

The thing is, I don't think we're near the boundary of power efficiency even with today's silicon. The way forward is to push all the feature tracking and other low-level vision tasks down to the ASIC level, while accepting some strict limitations on the rendering side. So a lot of it comes down to scope of features. Its most immediately useful applications are little more than a glorified HUD -- project some GPS data here or an info overlay there. Blit a video source on top of a marker, that sort of thing.

This sort of practical product, by 2027, I'll say yes. Mindblowing, transform-your-world stuff? No. But let's see what happens...