On-Chain, the Emperor Wears No Clothes

It's a pretty useful distinction. Additionally, a lot of the communication differences that drive engineers nuts (eg. selling a project as nearly done when actually it needs significant work, representing that there's widespread consensus among stakeholders when in fact there is no such consensus) are examples of epistemic vs. deontic reality. A manager's responsibility is to make what should be actually come about, and so their job includes moving people - sometimes with lies - such that reality changes in the desired way. An engineer's responsibility is dealing with reality.

Ideally there's enough trust between manager and engineer that both parties can bend and the goals can become something actually achievable, while reality becomes something desirable to the organization.

The way for engineers to reclaim epistemic power, though, is to realize that it's not your responsibility to protect upper management from the consequences of their decisions. If you trust management to be rational, sure, up-level your concerns about reality and hope that they adjust their "should". But if they still insist on something that you know is unworkable, go quit and work on a project where you can actually achieve something useful. Either you'll be replaced and your replacement can achieve management's goals (in which case, more power to them) or management will be forced to inhabit epistemic reality.