What's holding back the current state of WebGL + VR?

Just out of curiousity. Is it the difficulty in rendering the game twice? Or difficulty in receiving hardware output from a vive/occulus? Is there some kind of device limitation?

I'm curious in part because rendering limitation doesn't make sense as long as devs are okay with lesser demanding games. ASSUMING that, then the only option left is not being able to receive hardware output from headsets and controllers fast enough? Is that REALLY the issue??

If anyone could reflect on this or who has personal experience that can answer for sure, feel free to contribute your responses!

Thanks in advance!
- Eliot