Look at the diameter of the actual glas element of the (front-facing) camera. There can be no active display over that area without blocking most of the light (and affecting what the camera captures by the light of the pixels of the display). At some point, you might be able put a large number of really tiny cameras that peek between screen pixels (though I'd still say the display will need to give up a pixel or two for each tiny camera) and do computational photography (ie, assembling an image from a large number of separate cameras).They have been working to put the cameras behind the screen. Personally, I think 2019 is a bit early and 2020 is more likely.