Why Colab When You Can Fight?
Obstructionists can always wrap themselves in the thin cloak of “misson,” but when you get down to it, its about the credit (and the associated budget). Nothing illustrates that more clearly than this story. Lather, rinse, and repeat it for pretty much any defense, law enforcement or intelligence mission.
With rare exception, people in bureaucracies will default to strategies and tactics that reward them the most. Its human nature: I’m going to respond most effectively to the metrics I am measured against. Collaboration isn’t something I’m rated on, so the people I could learn from, or teach, or work with to put together a better case, can go scr3w. You don’t have to change the structure, mission or people: you just need to add a “play well with others” bullet on the rating form. Watch what happens when promotions and bonuses are tied to how actively and meaningfully you engage in joint endeavors.

You had me going until adding something about engagement to the rating form. We actually have this in my agency, but the problem is what exactly is engagement and when is it necessary? The fact is, it isn’t always necessary, but it is necessary overall. I was rated “minimally successful” once because my manager didn’t believe I was “going to enough meetings” thus I was not collaborating. I got that rating overturned (long story), but it speaks volumes in my opinion to why this is a dangerous idea. How about instead, we rate performance on “achieving results” and if collaboration is required to do so, it will be done or the product will suffer and the person should be rated according to the bad result, if collaboration is not required, the time will be spent elsewhere. In this case, your solution only further entrenches the bureaucracies that impede mission success every day.