See URL for the relevant thread on the WebGL mailing list. Quote from the first message in that tread: "One of the issues with WebGL in particular on mobile is because WebGL has to be composited it loses a lot of performance. What I mean is WebGL renders to a texture. That texture is then drawn into the backbuffer in the correct order with other HTML elements. This effectively means a fullscreen WebGL app is drawing an entire extra screen worth of pixels vs a native app." So once a canvas element goes fullscreen, it is expected that we can get quite the performance boost and framerate improvement if we can bypass the Gecko layout pipeline - the compositor - and write to the (GL) backbuffer directly. This was also discussed on the dev-platform mailing list, starting here: https://lists.mozilla.org/pipermail/dev-platform/2015-January/008277.html

Thank you for bringing this discussion here too, we'll be following this one closely as well! Again, if you ever need a real-life example of such usage or potentially a testbed game, feel free to send me a mail and we'll give you access to Project Atlas' closed alpha.

Yep, this is the plan. It'll take a little work, but it'd be really nice for a number of reasons. Hopefully something for Q2.

I can also provide large numbers of WebVR test cases.

I am removing the [webvr] tag as a VR specific rendering path that bypasses the compositor has been implemented as part of the WebVR 1.0 support in Bug 1250244. Perhaps the VR specific implementation may be a good reference for someone implementing the general purpose WebGL case.