Better to use discovered languages as they say, but also better to learn the actual abstractions of computation itself, which CT does a great job at, instead of arbitrary patterns that are based on human intuition.
What about the Curry-Howard-Lambek correspondence? This is directly relate to why math works so well on the real world. CT and pure FP languages seem to tap into this fundamental workings of our mind and possibly the universe (CT is used in quantum mechanics). In contrast things like OOP are modeled based on intuitive patterns that solve specific problems, and which are often based on metaphors or abstract analogies, but not a fundamental abstraction of nature. In other words CT seems to capture the patterns of computation itself.
