I was interested, though, in this:
Now, it’s one thing to collaterally maim and kill foreign civilians in necessary defense against an unjust attack by their country. But it’s another thing to collaterally maim and kill foreign civilians to whom we owe compensation! The second is much worse.
But, of course, the obvious response to this is "No, it's not." Once you're at the point of maiming and killing, whether you are in debt to the people you are maiming and killing is at best a secondary matter. And if you were to ask people whether your debt to someone might possibly be canceled if they attacked you without cause, there are a lot of people who would say that, yes, depending on the circumstances, that could very well wipe the debt clean. They may have already taken out their compensation in violence and then some. (Part of the problem, too, may be that compensatory debts for war are not actually equivalent to what they are compensating for; what happened was that lives were lost and infrastructure destroyed and time wasted, none of which can be undone at all. Compensation in this sort of case is to help start some things anew; it is not a measure of harms.) And the problem is seen in focus when we consider his further comment:
We end up, then, in an interesting place: sometimes, in order to wage a defensive war permissibly, we first have to discharge compensatory duties.
This starts looking very much like a reductio ad absurdum of the entire idea.
The sort of thing Bazargan-Forward is talking about arises from the fact that the principle of proportionality, as it is usually understood in analytic versions of just war theory, is a foreign principle to the theory itself, a consequentialist idea that wandered in and pushed out the original notion of proportionality, which was not about balancing harms and benefits but about fitting appropriate means to the acceptable ends. (That's where the name came from: proportioning means to ends. Thus you'd originally have considered things like what is required actually to solve the relevant moral problem, consistency of means with ends, the relative importance of various priorities, what you have a natural right to do, and the like.) Thus there's nothing about just war theory itself that says anything about how to handle it, allowing for looser and stricter interpretations as people please. Bazargan-Forward takes a very strict interpretation. He is thinking that we have accumulating harms and benefits that just stay there inertly, so that if you are in arrears it always has to be made up as soon as possible, and that you are not acting proportionately if you do not pay back what you owe.
But none of these things are true. As time goes on, what is harming and what is benefiting are constantly changing, and if you are actually weighing harms and benefits, that would therefore change, too. Before Petroland unjustly attacked Imperioland, Imperioland was indebted to Petroland because of past harms. This would not be static in the meantime as the populations changed, even considering downstream effects, but even if we set that aside, the unjust attack is now a harm in the other direction, and thus the entire calculation of who harmed who and how much would now be different. Further, hostilities arguably put the whole matter on hold as something that cannot be properly worked out until other problems are solved. And whether or not you are in debt simply does not affect whether or not you are engaging in a proportional response in this particular moment. If a criminal owes me medical expenses for prior assault, he still has the right to defend himself if I am trying to stab him in the chest, and he can defend himself in exactly the same way that he morally could if he were a perfect stranger.
This ties in to a broader issue, namely, with claims that just war theory can't handle this or can't handle that. Just war theory by its very definition is not a standalone theory. The root of just war theory is the question, "Can actions of war be just actions?" and the Yes answer comes with the condition, "when justly disposed people justly use means for just ends". The particular form of just war criteria is often affected by long experience of recurring problems in warfare and the kinds of things you can get people to agree to in order to resolve them (and sometimes by historical accidents like the consequentializing of proportionality, which is due to, first, Vitoria's particular formulation of it, and, then, a consequentialist interpretation, including the conflation of being harmed and being wronged, thus losing the sense of making sure harms you cause are not worse than the wrong committed against you), but all of the criteria are specific forms, adapted to the conditions of war, of general requirements of justice in any kind of endeavor. Any just war theory that does not implicitly recognize this is a fake just war theory -- why is it even talking about 'just war' if it is not about justice in warring? The question is therefore, can the relevant account of justice handle the problem? If it can't, no amount of fiddling with just war theory is going to improve matters, and if it can, there's no problem for just war theory.