This is the thing that has been true since way back in the demoscene from the 80s.
- if you can fake it fast enough you can get away with really low resolution/level of detail or inaccuracy.
i.e., in a game like competitive FPS, frame rate is everything. and in a game like that where frame rate is everything, nobody gives a shit if you can see reflections in the other guy’s eyes or whatever if it costs 50-70% of your frame rate.
Don’t get me wrong, ray tracing is the future, eventually. But we are WAY off yet. Until we’re surpassing say 4k144p with ray tracing on, heaps of people will turn it off for more frames. And even then, decent VR will require higher resolution than that, ideally at similar or higher rate.
I think one of the graphics guys from AMD (in an interview from a year or two back) was suggesting something like 16k res per eye was what you need for proper VR (with a decent field of vision). We’re simply nowhere near that, and rasterizing can “fake it” well enough that the frame rate difference is simply not worth it to do ray tracing yet.