Category theory is the lowest level of functional programming. At this low level you can do some pretty ground-breaking stuff (like the Haskell z3cat[1] library which can apply a SAT solver to a regular Haskell function), but it’s not required, in any way, in order to write great Haskell code, and make use of Haskell as a tool to write applications with very few runtime exceptions.
a) know why some things are done the way they are (in case you're not satisfied with "this works!"), or
b) if you want to find inspiration to write a certain kind of library: Haskell is one of the few languages that has, as I like to call it, "over 50 years of documentation".
--
To elaborate: a category is a bunch of objects, and functions (morphisms) between the objects. The types in a language like Haskell form a category (called Hask in this case) where the objects are the types and morphisms from one type to another are the functions that go from the former to the latter.
A lot of things done in Haskell are instances of general constructions from category theory applied to Hask. Comfort with these general constructions certainly helps to illuminate many ideas.
For example, the great trifecta of Haskell abstractions comprises three interfaces ("typeclasses") called Functor, Applicative, and Monad, which are used everywhere. Each of these corresponds to a concept from category theory: for instance, what a Haskeller calls a "Functor" is what is categorically called an "endofunctor on Hask". Among users of the language who know any category theory at all, most have some familiarity with things at the level of what I've just said, but not much more. A knowledge of category theory at any higher level (like in the linked repo) is not at all common, statistically.
Also, not many of the common libraries are based on deep categorical ideas (lens [0] is an important exception) but even then, knowledge of the mathematics is useful if you want to dig into how the ideas are implemented, not if you're happy just using them. Learn as little or as much as you'd like.
CT is great for "looking to get really good at" FP. It writes out a parallel universe where certain things are easier and certain things are harder but the same core pillars of FP are intact. In this parallel universe there are myriad examples of constructions, techniques, and conceptual frameworks which can be "ported" back to FP. Likewise, some FP concepts have interesting "ported" versions in CT.
Learn CT when you want to build intuition for how to construct functional architecture. Don't learn it to start and don't even learn it when you run into Functor/Applicative/Monad. It's _not_ useful for understanding them in practical terms.
Go learn it when you start composing higher level constructions together comfortably and start running into questions about the most efficient ways to lay things out. It's _great_ for that.
Is category theory worth studying if I (coming from c++ background) am looking to get into functional languages like Haskell? How common is it actually among Haskell or Lisp programmers to know and utilize concepts from category theory in their programming?