I feel it depends on a lot of factors, such as how senior your ICs are, how interlinked their tasks are, how complex the domain is, how large the project is, wh

I feel it depends on a lot of factors, such as how senior your ICs are, how interlinked their tasks are, how complex the domain is, how large the project is, what the rate of attrition is, how complex the organization’s established procedures are… An organization that has nothing but fairly senior researchers working on separate projects might need little management. An organization with lots of more junior engineers working on something like a rocket (where every part depends on every other part) might need more overhead.