I feel your pain: after bouncing off Haskell many times, I found that Elm was a great entry into a more practical and narrow way of experimenting with the benefits. Now I'm reading through the new Idris book, which has the same practical approach to more complex (to me) concepts.