The Makhnovshchina was a guerrilla formation, an army of sorts, which was one of the reasons that anarchists in other parts of Russia were skeptical of it as an *anarchist* guerrilla formation. The fact that there was a de facto officer corps, that there was a quasi-militarist style of discipline, and that they entered into treaties and/or cease-fires with other forces that were clearly militarist (the Ukrainian nationalists under Petliura and the Red Army), all create a strong impression that their anarchic characteristics were mainly exhibited away from the battlefield.
While self-organized at the outset of the Spanish Civil War, almost as soon as the anarchist militia columns began to accept the strategies of their military advisers they began to lose their anarchist characteristics. For the sake of efficiency and strategic parity with their enemies, they began to instill more and more militaristic discipline around taking orders and coordinating attacks with their (unreliable) allies on the front.
So what precisely are you trying to indicate with these examples? That anarchists fight in wars? To the extent that's true, I would argue that they lose their anarchist characteristics. In other words, their anarchist credentials exist in inverse proportion to the extent that they accept militaristic logic.
In terms of the other types of conflicts you list, beyond a certain number of participants (for the sake of expedience, I'll call it "tribal"), the ability to muster the combatants and support infrastructure required to maintain their fitness for battle requires the logistical resources of some kind of apparatus that can requisition (by force if necessary) men and material. That is, a state.