It’s as far as I know only used in the Julia parser (and easteregg julia --lisp not even documented with --help-hidden). I don’t think it’s going away in the any time soon (since I read, it’s not "speed-critical), but it’s been discussed getting rid of FemtoLisp, there’s already a parser for Julia written in Julia-only code, but I’m not sure it’s always in sync, since Unicode symbols occupationally get added to the parser.
And a different parser (currently done in C mostly I believe):