It is rarely within reach to draw new insights from applied category theory, in particular because of the Yoneda lemma and the greater familiarity of sets and functions, and also because as algebraic objects categories have very few properties.
Bare categories do not give much insight in pure mathematics either, it's just a common language; interesting things are categories with lots of extra structure like toposes, derived categories, infinity-categories, and so on.