Your analogy doesn't work. A "product" in research pure math is not the same as writing code.
You can import a theorem that another paper proves without having to understand the mechanics of the proof itself.
You can import a theorem that another paper proves without having to understand the mechanics of the proof itself.