Whenever I try to be productive in Haskell I end up taking the research phase too far and end up over-my-head in category theory that I don't understand. I'm never productive in Haskell.
Haskell allows very "deep" abstraction - but what I find is that you should not overuse it. Abstract where practically useful, otherwise don't go for it. :-)
You might not be productive in the traditional sense of "delivering business value". However over the years of letting Haskell change my brain a little piece at a time, the results have accumulated. I wouldn't swap that deeper understanding for all the productivity in the world. :-)
Haskell is a great language for delivering business value. The typed FP paradigm encourages reusable code much more than any other paradigm I have tried.
Yes, absolutely agree with you. I wish I was
writing Haskell day to day.
My point was that it takes a long trivial amount of time to learn Haskell during which you might feel "unproductive" by that measure. I feel that it's during that period that the magic happens :-)
Start by coding. Be prepared to rewrite your first big project once or twice (good thing is that you'll realize it needs a rewrite sooner, rather than later - anyway, don't start with something too big), but start coding. You won't learn it by reading.