I am glad to see this and hope it all turns out to be good information, not just more sales hype. On ASTRAEA civilian autonomous flight program that I worked on in the past (pamp pamp! trumpet blown), it was established for their purposes that vision is enough for any air scenario including even air to air refueling, which is a highly safety critical activity, taxing even the best manual pilots to the edges of their skills do reliably every time.
There was a whole stack of other sensing techniques including Lidar thrown out, after it was realised that a machine only needs to be equipped with vision at least as good as that of a human, to do at least as good a job of flying, given also similar requirement on acceleraration sensing (the human just "feels" it).
So it was to be expected Musk / tesla's work on that would pay back in the end.
So it's good he had enough confidence and sense to stick by the decision to ditch the other kinds of visual sensor including Lidar.
But I still say he's dead wrong about hydrogen.