Your analysis could be applied to Creo without much change because you're looking at implicit incentives in any place managers (or resource-controllers) are expected to accomplish goals.
That said, I think your analysis is a bit weakened by a moral elements (good managers vs bad managers). And even more, it's not necessarily bad for an organization to have locally controlled resources. One can posit an omniscient central controller that would manage resources better than local managers but it's likely that there are limits to both central controllers and local managers. What's "best" depends on the task, the organization and who's judging the result.
That said, I think your analysis is a bit weakened by a moral elements (good managers vs bad managers). And even more, it's not necessarily bad for an organization to have locally controlled resources. One can posit an omniscient central controller that would manage resources better than local managers but it's likely that there are limits to both central controllers and local managers. What's "best" depends on the task, the organization and who's judging the result.