It occurs to me that the central problem with corporations is one of causation. No one quite understands the chains of causation from effort to output, product to pollution (or any unintended effect), and the link from management attention to change in state or results.
It's all a mystery, really. This is one of the more important domains in which systems thinking and dynamic modeling can improve things. It is difficult to encourage autonomy when people are unclear about what to do, about how much Jeremy's efforts might have helped and how much Jessica's hindered, or vice versa. For now, the corporation is a big black box and one proof of that is how strictly defined are efforts and goals within corporations. It's rare that managers are working with a testable hypothesis about how to improve and even rarer that they are collecting data to help decode the system dynamics they presumably manage.
Advances in data collection and computer modeling may allow for sophistication in simulation and modeling that exceeds that of which a human manager is capable. Causalities could be revealed in much the same way as those found in pharmaceutical testing or meteorological models. Brains are simply not big enough to simultaneously compute the wide variety of causes that play into outcomes. In less than a decade, we'll find managers working without tools for modeling systems dynamics as quaint as we now find ditch diggers working without backhoes.
As people within corporations and stakeholders within communities better understand the causation of complex corporations, resources, time, and attention will be better applied to more profitable and less unintentionally hurtful outcomes. Without such tools and models, we increasingly find ourselves trying to steer from the trunk, tasked with managing complexity we're unable to comprehend.