for what it's worth, back when i was in b-schoo, the predominant theory for the existence of large, diversified Asian conglomerates (Japanese keiretsu is a prime example, but it's not unique to Japan. Similar entities exist in S.Korea and Thailand, among other places) is the friction of starting and doing businesses in these countries. Startups there have such a hard time getting funded and landing clients, that the financing aspect alone give the conglomerates enough of an advantage to edge out startups, despite the inefficiencies typical of large organizations (and in this the Japanese corporate giants are no better than IBM or GE).

this theory is typically mentioned when introducing the Theory of the Firm, i.e. why do companies exist at all? why not everyone just freelances, and when you need a marketing/finance/legal/coding person, you just contract one on the free market? the idea is that there are always frictions when doing business with someone new (is this person good? trust worthy? how do i find out?), and how much friction there is determines how big the firm will grow (to incorporate functions in-house, or expand to other industries).

in a perfect frictionless economy, it could indeed be true that business transactions all happen at the smallest unit, i.e. the individual. at the other extreme, there would just be one firm that coordinates all economic activities. all real world economies sit some where in between.