One of the people tasked with deciding this seems to have no concept of integration costs. He thinks that we just grow our team by 2x and then we're 2x more productive. Even though their technology stack is on JVM languages, and ours is Python-based.
My experience is that this is totally underestimating the costs of integration, but are there studies to back this up?
(Let's set aside the coordination costs of a larger team -- we can refer to the Mythical Man-Month for that).