One should distinguish between short, trivial and elementary proofs. Trivial proofs can be omitted with no harm. Omission of proof in the other two classes does seldom any good *to the reader*. In you shoes I would put full proof in the preprint on arXiv. It can be easily edited out in the published version.