Note: this question concerns the reason I started posting on this SE years ago, and has to do with my obsession with universal sets, anti-terms (antisets of late), the morality of punishment and destruction, etc. so basically, the majority of my questions on this site. Not, that is, that deciding to limit the use of the "all and only" quantifier, say, or trying to develop weird variants of epistemic logic, or whatever, are directly what I'll be on about, but just the same... so I'll divide the question into two main parts, in such a way that if the background reasoning doesn't seem relevant, it'll be presented after the question itself is posted.
The actual question: so, in the Religion, Kant says that the Stoics were mistaken to think that inclinations (very roughly, physical desires) are our "moral enemy." So he goes on to make all sorts of somewhat random-seeming comments about demonic spirits, the Antichrist, self-crucifixion, levels of radical evil, etc. But so he still seems to portray analogies and allegories of spiritual warfare as adequate to representing inner moral tension/conflict.
How is this realistic? At least a physical desire, like competitive anger or extreme hunger, is not only tangible but "displays" in a "fightworthy" manner, and not just in the sense of "providing us with testing grounds for virtue." But the non-image of diabolical will is more than abstract or generalized; it is a void in our minds. It already seems as if layering our propensities (animality, humanity, personality) and splitting transcendental freedom into Wille and Willkür is perilously close to making determinate claims about how transcendental freedom is structured and operates. But even these structures/functions don't seem to lend themselves to combat tropes. It is entirely possible, and even often probable, that I am tempted to do my worst when I am least conflicted, emotionally, about what I am doing. Putting myself in some sort of "fight mode" vs. that state of mind isn't just silly-sounding, but how could it even happen? It will be exactly because I'm not in self-directed "fighting stance" that I won't fight!
Consider, for example, what Kant says about the potential, and even sometimes actual, moral value of friendship. He says that it exists in some apparently perfect form, in the physical world, at times. I didn't understand how this phenomenon could be set apart from complete redemption until I considered that a person can have a sufficient moral relationship with whosoever they think is morally relevant, but they can at the same time have an immoral, and hence deficient, standard for judging who is included in the sphere of moral relevance. Kant at his most racist is a good example of this situation, as are racists broadly. Even so, then, what is there in the redemptive influence (not total effect) of pure friendship that corresponds to combat imagery? Do I represent the good side of my friend as fighting the evil side of myself, and vice versa? I'm not saying, "Yes," or, "No," to this question; I really don't have a good enough answer right now, and I'm actually at a point in my life where figuring out whether to represent myself as having to face someone in terms of that person as my moral enemy, is very acute. This is someone I did not think I was supposed to (more or less literally) fight, but now I wonder if that's exactly what I'm supposed to do.
Maybe it's not actually so important to Kant to talk the fighting talk, and maybe it wouldn't have ended up seeming like an issue to me, but I am reminded of some ominous remarks Cornelius van Til, an extremely fanatical Reformed epistemologist (and he was, genuinely, an epistemologist, not just a theologian), made about some other Reformed scholar criticizing him (van Til) for using military imagery to talk about the opposition between Christ and the saints on the one hand, and Satan and the so-far impenitent on the other. Van Til's critic said that using these metaphors ran the risk of encouraging subconsciously violent perspectives, and van Til's response was to not take the problem seriously.
And so yet fast-forward just some few decades (less than a century, anyway), and we have this wacky cult infecting half the Internet, and a lot of countries besides, largely founded on not just the kind of logic that van Til championed, but his actual writings on his theory of logic specifically. (By the way, my friend-or-enemy-or-whatever was heavily involved with this cult, and still might be; that's part of why I wonder about having to fight him...) They portray themselves as fighting an evil world order, as well as evil aliens and evil spirits (when they don't claim that the aliens are the spirits). Their telos is the day when a storm of truth and death washes across the Earth, wiping out the "lies" of the imaginary world order and the lives of the imagined leaders and agents of that order. If you're familiar with some of their more rabid posts here and there (everywhere, sometimes!), I think you could recognize how their mindset seems based on an idealization, on multiple levels, of combat metaphors for moral progress, both on personal and social levels.
What hath Auschwitz to do with Jerusalem? (This is the tangential section, included just to show what I think is at stake in this debate about (im)moral combat imagery.) And so I think the above really is a problem. This is my reasoning: back in 2016 I was outlining the endgame for a convoluted story I wanted to write down as a series of books or something eventually, and I made part of the plot turn on there being two, not one, Forms of Evil (to represent evil's will to fighting as essential, so that the essence of evil has to itself divide so that evil can fight itself), except that the other evil Form managed to deceive itself and most everyone else into calling it just "the Form of Destruction," or to be fancy "Apollyon," instead. There's also an angel (basically) who betrayed his God long ago, but then genuinely repented and helped stop a terrible war by using the so-called Form of Evil's power to seal away the Form of Destruction, but who then makes an even more elaborate mistake in moral judgment by deciding to not only awaken Destruction later, but to do so for the sake of using Destruction to destroy the so-called Form of Evil. The angel thinks that this will redeem Apollyon, and that if any being just has to be sacrificed eventually, it's got to be the Form of Evil. Instead, Apollyon turns the Form of Evil into the Form of Nonexistence (to metaphysically realize the old doctrine of evil-as-privation), except as a mostly mindless will to annihilation, the Form of Destruction then negates the Form of Nonexistence, which having been merged with the Form of Evil then is recreated so that all previously nonexistent evil now exists. Then Apollyon kills Evil again, and so on and on, back and forth, and there's a rule in the background of the story that says that killing Apollyon will unleash so much energy that the entire multiverse will collapse, so the world is caught between the constant mass recreation of all other evil, and the one other final evil of multiversal death.
The point of this, as a conceptual exercise, is to express why the thought of destroying evil, even destroying evil forces of destruction, is beyond misguided, ultimately: because this is what that would lead to, if carried to its metaphysical conclusion. But then the solution is not killing Apollyon, not killing Evil, not punishing them in some other "gruesome" manner, but is not supposed to be violent at all. In fact, as far as the story goes, the solution involves everyone else in the multiverse giving all their spiritual power to Apollyon, so that its will to negation inverts upon itself to absolute infinity and becomes a nonnegative force. Given the fantasy nature of the plot, the Form of Evil to the side then does have its own will, and so its own penitence is possible, and though I never conclusively imagined how that option would be pursued at the end of the sequence, it would have to be pursued (seeing as the other option had to be rejected for the sake of the story not ending on a bad note).
The question seems similar to the debate over violence in video games, or "whether one can do wrong in a dream." There was a comedy video that I watched some months back where they show a stereotypically devout mom objecting to her son playing a violent game until he tells her that you fight demons in the game, whereupon the mother assumes that the play is somehow expressive of service unto Christ. (This was not quite my own experience: many, many years ago, long before this comedy skit was done up online, I asked one of my grandparents to buy me what was, I think, either the same exact game they used in the skit, or a game in the same exact series (90's Doom), and I tried to justify my request on the same ground that the son does in the skit. But my grandfather, though he did buy me the game, was not really swayed by that argument, and in fact I don't know what swayed him at all.) So, one way to approach answering this OP question objectively and not in an opinion-based way would be to cite research into correlations, causal or not, between consumption of violent media in general and performance of real-world violent acts; then to fine-grain the citation by seeing whether portraying said violence as justified is a better indicator of consequent aggressive behavior.
But so I think this might better be it, for me, for now, here. I love analyzing set theory, it's turned out to be one of the most fun things for me ever, but I don't think trying to come up with new weird questions about the philosophy of mathematics is going to go anywhere right now. As I said, my main problem right now is deciding whether this one man, to whom I committed my whole theory of romantic idealism (as based on degrees of infinity in set theory, no less), really is my objective enemy on account of his wrongdoing, or whether I have to deal with what he has done in some less confrontational fashion.