It seems to me that MetaOcaml is an extremely valuable tool with many interesting use cases would it be incorporated into the core OCaml distribution.
AFAIU, the current BER implementation is not very invasive, are there some plans indeed to merge the BER implementation into the mainline?
Quoting @gasche from the comment section of a recent reddit post:
Are there any plains on mainlining it in the future?
To my knowledge Oleg has no plans to stop maintaining it. It is actually a relatively small patch on top of the upstream OCaml implementation. (We discussed upstreaming it, but then Oleg said that he wasn’t sure the design was final, and would prefer to wait and gather more experience/reports before considering upstreaming.)
“my 2 cents” :-).