Your comment is valid if you assume that a business works like one homogeneous organism, in the language of biology, you have one organism on which natural selection (in this case, the market) acts.
I think, especially with larger companies, you have to treat the whole organism as a bunch of sub-organisms that works together (most of the time) towards a common goal. In that case you can have several different forces acting, for example, the size of a department may be a positive attribute for the manager of that department, even though that size drags the entire 'super-organism' down. The manager probably also goes out of their way to 'protect their people', again to the minor detriment of the super-organism. But is that bad for survival of the super-organism? Probably only minor!
Think of behemoths like IBM - think of how many international dependencies they have, think of how many sub-departments they have, think about how often these departments must fight with each other to make themselves look good to the higher-ups. Think about how much money they make and continue to make, it takes a lot of infighting to get such a big company down.
The way I typically understand this is by Dunbar’s number: that 150 or so persons who are “real” in any given person’s head. Any org larger than ~75 has this type of multipolar dynamic where different sub-units are pulling different ways.
It’s held together by the myth of money and/or organizational purpose, which is actually less immediately “real” than any one of those sub unit in-group people.
It’s not that explicit, but it is certainly innate: large orgs all suffer the same way.
True it takes a long time for formerly great organizations to die. I think a lot of the problems with this type of analysis are time scales and costs of switching. IBM is probably not the best solution for anything if you're starting something new; but if you already have something it's probably good enough and it would cost too much to change right now.
So IBM can limp along until new organizations develop products that are sufficiently cheaper or better that it makes switching the logical move. Or the companies tied to IBM eventually fail and replaced by more nimble organizations.
I think, especially with larger companies, you have to treat the whole organism as a bunch of sub-organisms that works together (most of the time) towards a common goal. In that case you can have several different forces acting, for example, the size of a department may be a positive attribute for the manager of that department, even though that size drags the entire 'super-organism' down. The manager probably also goes out of their way to 'protect their people', again to the minor detriment of the super-organism. But is that bad for survival of the super-organism? Probably only minor!
Think of behemoths like IBM - think of how many international dependencies they have, think of how many sub-departments they have, think about how often these departments must fight with each other to make themselves look good to the higher-ups. Think about how much money they make and continue to make, it takes a lot of infighting to get such a big company down.