Looks solid to me and I saw a presentation from @ntBre this morning using this experimental functionality, so I assume there aren't glaring errors. IIRC, this is the first time that we're introducing an indexed parameter that could be int or string. I'd approve this pending adding tests that ensure setters and getters do sane things, including being set to int/string while being accessed both as parameter.idivf[N] and parameter.idivfN.