That’s a wildly impractical position to take and isn’t really true in my experience. I spent several years at a seed stage startup where it was me on backend an

That’s a wildly impractical position to take and isn’t really true in my experience. I spent several years at a seed stage startup where it was me on backend and one other engineer on mobile. We would specify a JSON API and then could work independently (and very fast) to realize the functionality because he’s an expert at mobile programming and I am an expert at backend. We didn’t have the luxury of working at 50% output for a year to learn each other’s specialization while trying to get the product off the ground. He didn’t need to learn the ins and outs of rate limiting with redis and I didn’t need to know how to animate a custom UI component for each of us to be maximally productive. Division of labor and the interfaces which enable it exist for a reason. I suspect it’s because it reduces the complexity of collaboration between engineers with different specializations. It facilitates the separation of concerns.