If I understand the argument in this article correctly, then a highly social animal such as a wolf, carries an expectation about the various relationships he forms with his peers. He is also able to understand a cost/benefit risk assessment in regards to a dominant so that it may not appear worth it to a submissive to jeopardize its safety in the pack and the possibility of scraps that remain after the dominant takes its full share. Furthermore, he is able to comprehend that over time the cost derived from being submissive can become greater than the risk of finally challenging a dominant and fighting it out. So if I have it correct that the above is possible in the mind of a wolf, then therefore it must also be apparent to a wolf that a fight with a peer can potentially threaten pack cohesion when hunting a large, dangerous prey animal. So if dominance and submission in wolves derives from such cognitive processing in peer-to-peer relationships, then therefore the most logical factor that mitigates violence and induces the many to become submissive to the one, would be for the purposes of maintaining group cohesion in the hunt so that the pack can present a unified and syncopated front in the face of a formidable prey. In other words the peer-to-peer relationships they depend on for success in the hunt, and upon which the survival of the entire pack depends, would be the true “efficiency” metric which mitigates internecine conflicts so that they don't escalate to dangerous levels. The group dynamic of the hunt would also better explain the social inhibition apparent even in the superior in so called dominant/submissive interactions. So whatever dynamic accounts for the coordinated action of the group in the hunt, would also account for the coordination of personalities in the pack.