[altruios]

I mean to say, modeling a company hyper accurately - down to personal psychology of the high level employees - has diminishing returns for predictive value, and also diminishing returns for generalizability. And a surprisingly useful toy model (the egregore and it's hand eating lifecycle), describe better(or at least, good enough) what will happen with any particular company given it's current state according to the terms of that model.

To clarify the possible misunderstanding: 'its' is referring to the toy model of the egregore (not any particular company/egregor), My point is useful things are useful regardless of them being strictly true.