He favors #Go, with #Rust as a second. Believes that !DLang's future was harmed by the Phobos vs Tango standard lib shootout a few years ago. Dismisses #Nim, #Ocaml for their lack of real-world usage. Believes that #CPlusPlus was a mistake that only got worse as it become more complex.
I wouldn't say #dlang's future is harmed so much as delayed. Andrei Alexandrescu has remarked that it seems to take 5 years for the bad impression from the split to go away (IIRC)). D2 (current version) does seem better designed and picking up stream but I don't see it as a viable mainstream language yet. In a few years however it can come a long way, especially the BetterC stuff.
Harmed? Delayed, maybe. D is growing faster than ever.
OCaml and F# are used in the financial industry. But they probably won't replace C because people are scared of no-parens almost as much as they are scared of parens. A nice mix of parens of various shapes keeps people calm. :-)