The article assumes an existing team which is a bit problematic when talking about time to market. If you start the analysis earlier (two people discussing some

The article assumes an existing team which is a bit problematic when talking about time to market. If you start the analysis earlier (two people discussing some ideas in a coffee shop) then I'd argue that a language like Haskell can be problematic if your metric is time to market. You might very well make that up later by having a more robust code base or reaping any of the other asserted benefits but the existing gallery of premade and tested building blocks in other languages seems to be richer. It's probably also going to be harder to add people to your team (on average). I would have liked to see a comparison to other functional languages (say Elixir or OCaml) and not just Java and C#. I'd also argue that picking Java instead of a more agile environment (there are some cool lightweight Java frameworks but most people will associate it with the rather heavy enterprise stack) when comparing regarding time to market is a bit odd. Granted I'm mostly thinking about webapps (but the article mentions Yesode). Still a nice article (since my post sounds overly negative upon rereading).