Even with humans, the first publication is rarely the best expression of the lesson. Getting published does more than establish priority; it also frees up the community to build on the result. Nobody expects a human to wait until the proof is comprehensible to anyone except themselves and the referees.