I vaguely recall reading a story about this report — the Haskell solution was considered quite elegant, but some people felt that it was somehow "cheating" and not a real, fully-baked, solution because of the way it used functions to encode regions.
I don't think that disqualifies it for enterprise. People will start as newbies and they'll have a very difficult initial hump to get over but once they've been using it for a few months they'll be productive and after a few years they'll be very productive. Unless turnover is so bad that a substantial portion of your workforce is always "new", use of Haskell should be a net positive for a company.
There was a study a while ago (http://www.cs.yale.edu/publications/techreports/tr1049.pdf) that had programmers experienced in each language solve a simplified but non-trivial problem for the US Navy in their given language. Haskell was rated quite highly. There was even a case where a newly-hired college graduate was given 8 days to learn Haskell, with assistance from an experienced Haskell programmer. The college graduate was then asked to solve the problem using Haskell and still performed very well.
So I don't think it's impossible to learn Haskell quickly either, it just requires access to an experienced mentor, something an enterprise could provide.