Chrome Version       : 66.0.3355.0

this test is a not a good guard against the regression from  issue 372314 . Since the document layer is detached we don't actually check that we correctly calculate the touch rects on the video controls. It appears that the media player has since been modified to not require touch handlers and the expectation reflects that. As such, this test can't tell the difference between the controls not having touch handlers and Blink failing to calculate touch handler rects.

The test should be replaced by one that actually tries to use the media player controls via simulated touches and confirms it affects the video.