His thesis is that none of the current crop of functional languages will rule, but that some new functional language not yet invented (or widely known) will.
A language can't come out of nowhere, because it won't have the critical mass of people, libraries, etc to be usable for "real work". IF a functional language is going to be mainstream in only 5 years, then there are a significant number of people using it today, and it will be generally known in the FP community.
Even F# was a) in beta for years b) solved the library problem by piggybacking on C# and c) had a critical mass of OCaml users who could jump straight into it.
