The author spends a lot of effort trying to make his functions point-free. For example, the author first writes:

    inexactMatches xs ys = matches xs ys - exactMatches xs ys

then decides it's not very "nice", and ends up with this:

    inexactMatches = pure2 (-) <**> matches <**> exactMatches

I can sympathize with this preference. When I first started learning Haskell, figuring out how to make something point-free was extremely satisfying. And it also increases DRY, which is generally a good thing.

But over time I realized I was overusing point-free style. It was fun to write, but harder to read. If I were actually writing "inexactMatches" for something real, I would strongly prefer the first over the second. From what I've read, I think many people take a similar path when learning Haskell.

That said, I think everyone has a different threshold. For example, I've heard many people say that a for loop is more clear than something like "map" or "fold". I would disagree with that, but maybe that's just because I've gotten used to "map" and "fold".

So maybe the author's brain is just able to handle point-free style better than mine can, but I do think this particular example if far enough on the spectrum that most Haskell programmers would prefer the first version (for real code).

So if you're considering learning Haskell, don't be scared away. You will have to start getting used to point-free style, but probably not to this extreme.