Often when I see ethical categories they fall into a rather neat binary of good and evil, right and wrong, moral and immoral, or some other duo. (I’ll just use “good” and “bad” from here on out for simplicity.) Sometimes the binary will be complicated by expanding into obligatory, permissible, forbidden, and superogatory. (That is, stuff you must do, stuff you can do, stuff you must not do, and stuff that’s really nice if you do but you don’t have to.)
You can also combine these two for eight total options (or, more likely seven actual options—I doubt there’s sense to be made of bad superogatory actions). The good obligatory actions and bad forbidden options are obvious. But then we also have necessary evils, that is, actions that are bad, but because every other option is worse, they are nonetheless the only acceptable option. And we have some actions that may be considered good at least in some significant aspects but because of some overriding factor are forbidden. For instance, one might believe that stealing to serve the needy is good but the law overrides the goodness. (One may argue that the overriding factor just makes it bad, but there is at least the conceptual space for the argument to be had.)
We also often see some understanding of things being more or less good or evil. While one might condemn both jaywalking and murder, murder is worse. This is already apparent in the above description of a necessary evil being the least bad option.
An element I don’t see played with as often is the location or even presence of the middle line, so to speak. Some theories even look bad because the dividing line between good and bad. For instance, if one takes a consequentialist theory to say that the only good action is that which maximizes whatever the good consequences are and every other action is bad, then the theory seems pretty ridiculous. There are plenty of good choices that could have been better. That particular ridiculousness is found only in the misplaced middle line. Perhaps the line is somewhere else, with a multitude of good and bad choices to make.
Or, more radically, maybe there is no middle line. Maybe for any two choices one can be better or worse (or of the same valence) as the other. This seems particularly intuitive to me because the goodness and badness of choices, outcomes, and everything else does seem to be relative to some sort of standpoint. Any neutral line seems like little more than arbitrary, especially if inaction is properly recognized as itself a choice.
The other aspect of value space I think we need to question more is its boundedness. I’ve encountered some people who think that there is a cap on how good things can be and everything short of that is badness. The opposite can exist as well, and I’m somewhat inclined to it: there’s an absolute value minimum, and everything is building goodness on top of that. Of course, there’s also space for having both maximum and minimum value as well as value being unbounded on either side. What domain is being modeled will make a difference. For finite choices, obviously there are bounds. For total states of affairs, the bounds are much less obvious, if they exist at all.