I think it's much different from a legal standpoint. First and foremost, the driver is assumed to be fully valid and thus in no need of a blind trust (no pun intended).
there isn't network delay when your passenger tells you its "clear right"
Yes; also, if the person is in the car, they're risking their own life -- not so if they're hundreds of miles away.
Not to mention the vastly better field of view and detail. Neither of which translate very well through a phone camera.
I have heard that if you were to, say, wave someone through at a stop sign and they got in an accident you could be held liable.