Honestly I think the biggest down side is that there's not enough commercial endeavors using Haskell, and thus theres horrifyingly few people working full time

Honestly I think the biggest down side is that there's not enough commercial endeavors using Haskell, and thus theres horrifyingly few people working full time on many core pieces of the ecosystem. Yes, my biggest critique is that all the great stuff in the Haskell ecosystem is the result of a small collection of smart folks helping out in their spare time. It makes me wondering what magic would happen when those folks can work on helping the ecosystem full time! I have to say that one of my favorite things currently about haskell is how nice and easy the c ffi is to use. So darn simple! (I'm also GSOC mentoring some work to provide a nice C++ ffi tool too). Theres so many great tools in the Haskell ecosystem, for every problem domain. Its not perfect, and theres always room for more improvement, but those improvements are happening, and the more people invest in supporting the community, the more those improvements happen! For example, one thing i'll be exploring in the neat future is how to do good Numa locality aware scheduling of parallel computation. It looks like i might be able to safely hack support in via a user land scheduler (though i'll find out once i get there). My principal work right now is building numerical computing / data analysis tools, and some of the things I'm doing now would be simply intractable in another language.