Since https://chromium-review.googlesource.com/c/chromium/src/+/853413 landed (and also while the pre-revert version was in), there has been a ~1.5ms increase in draw_frame_wall_avg during the VR browsing Telemetry test as seen here https://chromeperf.appspot.com/group_report?keys=agxzfmNocm9tZXBlcmZyFAsSB0Fub21hbHkYgICQmN2TnQgM.

This doesn't appear to be related to the background at all, as the test still uses the default grey plane/grid, and no snackbar shows up asking to apply a downloaded component.

So, it looks like some code change in that CL is causing an unintended performance regression.