I took a look at Elm a couple of years ago and while I was intrigued I simply couldn't consider using Elm in any serious endeavor. Not because it isn't capable, but because the risk of it dying if/when Evan stops working on it (for any myriad of reasons.) Yes someone could fork it, but that doesn't ensure it would continue to be well-supported.

Remember RethinkDB? When they shut down, the software was open-sourced to much fan-fare, but for all intents and purposes it appears progress has come to a screeching halt: https://github.com/rethinkdb/rethinkdb/graphs/contributors

Elm is unfit for any commercial software project IMO as long as it remains a one-man show. I (like everyone else) have limited amount of time to learn new tech, and despite Elm's technical merits - it's not worth sinking any time/effort into it at this time when other more viable options exist.