The URL proves that for the certain problem that set of people could come up with the best solution in Haskell. This might just mean that whoever was programming in ADA wasn't as good. The sample size is so small that makes it hard to accept it as evidence. Also almost everybody got A in the evaluation, if the others were C or B- I would see it a little bit more relevant.

I am with you, I also think that functional languages the way to go and I almost exclusively using those, but I would not make this big of a jump.