I think this is akin to the general problem people face when trying to convey why Haskell (or a Haskell-like library) is useful/awesome to them to a target audience that isn't already familiar with the subject. Simple and straightforward examples struggle to show any benefits that generalizations could/would provide. Complex examples are hard to explain without loosing most of the audience in terminology they don't understand. Or they just go on for way too long because they constantly drift off into explanations that don't really move the overarching subject forward.