Yeah, it just needs to work on its tooling and library story. Last time I tried, the OCaml toolchain was such a big pain that the Reason community basically told me to give up and compile to Node. And then there are relatively few native-ReasonML libraries, so you have to figure out how to integrate with OCaml which usually means learning how to _read_ OCaml without clawing your eyes out (I'm only sort of joking).
I'm rooting for Reason, but it has a few nontrivial hills to climb before it's practical.
That's encouraging to hear. Unfortunately I don't do much frontend web development, but hopefully if the frontend web-dev story catches on then the backend will follow.