I think with the right, bright people, they'll train themselves in no time. But sure, that shifts the problem from getting "people who know F#" to "people who are very intelligent and experienced", which might be equally hard if not more so.

I believe it's worth it, but have no data to back it up.