There is currently no good way of translation a position in a browser to a position on the screen. Our particular need is translating a gaze-position (which we have in screen coordinates from our eyetracking hardware).
It is possible in Firefox, using the proprietary mozInnerScreenX/mozInnerScreenY, but not in general. One of the comments in https://bugs.chromium.org/p/chromium/issues/detail?id=151983 seems to indicate that the correct course of action is to raise the issue here, so I will just post this and see where it leads. -- /Jan Norden