I think the use of "management science" in the title is problematic as management science is normally defined operations research or industrial engineering. Man

I think the use of "management science" in the title is problematic as management science is normally defined operations research or industrial engineering. Many of the techniques that you espouse, such as A/B testing, are examples of "management science." I think what you are taking aim at that's "non-transferrable general management" is the bureaucracy of large organizations. Part of it goes back to the "startup dollhouse fantasy" which is a term you coined that I really like. But I think what you are ridiculing with it is the imposition of formal control structures on teams that are small enough that social process and peer pressure is an adequate substitute. In "Corporation Man" Anthony Jay talks about 'hunting groups' or 'ten groups' having different rules from 'the camp' or the full tribe. Startups are like hunting groups, but every corporation has many "hunting groups." If you take a manager of a hunting group and put them in charge of a startup they will be right at home. General managers run the camp or the tribe and are less suitable for the special needs of the hunt.