I'm retired so I probably do not have access to the required separation standards; however,
Camera images are not visual separation as defined by any manuals I have access to, the usefulness of a control tower is (was) all about visual separation.
Radar separation between successive aircraft is 3NM, typically visual separation can accommodate close approaches, successive departures on diverging tracks, and go-arounds until a radar standard can be applied.
Eyeball separation has no tolerances applied to the image in the controller's brain. The eye is presumed to have an uninterrupted power supply, is backed up by the eyes of other tower controllers, has an advanced display system called the retina, and an advanced intelligence system attached that can instantly detect a subtle deviation from the cleared flightpath or taxiway. What tolerances will be applied to artificially generated images shown to a controller on a television screen? The new services will also be classed as screen-based, hence subject to the screen break requirements for the controllers = more controllers.
I love the identification capabilities of the camera system, but, once again, does the aircraft have to squawk ident to ensure the label is correct?
So many questions.......