1. The Sifri darshens that when Bnei Yisrael waged war on Midyan they were told to not surround the Midyanim, but to leave one side open for escape. A surrounded enemy is a hopeless enemy and a far more dangerous enemy – the Torah doesn’t want us to wage war this way. Ramban counts the command to fight in this way, “Vayitzbe’u al Midyan,” as separate mitzvas aseh; Rambam does not. Meshech Chochma explains that the Rambam held that this din is just a detail that falls under the broader mitzvah category of waging milchemes mitzvah.
Yesterday we discussed the chiddush of the Rogatchover that the battle against Midyan was not technically a milchama, but rather was an act of nekama. Perhaps the machlokes Rambam/Ramban hinges on that chakira. Rambam's view (as understood by the Meshech Chocma) makes sense if one assumes the battle with Midyan was a war like any other war, and therefore falls under the umbrella of milchemes mitzvah. Ramban may have disagreed because he understood the fight with Midyan as nekama, a distinct and separate category from normal warfare.
(It is still a bit difficult according to Ramban why the mitzvah given in the context of nekama should apply to other war situations as well. It seems Ramban understood that the mitzvah relates to the act of fighting, irrespective of whether that act takes place in the context of nekama or milchama.)
2. Rav Shteinman in his sefer on chumash raises the question (link) of whether a bracha was recited over killing the men of Midyan (or over killing Amalek). It sounds from his wording that his safeik was only with regard to the killing of Midyanites in particular, which he compares to misas Beis Din, but not to war in general. Again this reinforces the notion that the parsha of nekama is a unique category.