To an extent, SDL "presupposes" agency, or presupposes its own relevance/applicability modulo agency, from what is called OB-NEC onward. That is the idea that OB¬(A & ¬A) holds. However, as the SEP article notes:
Now consider the following statement:
(1) Nothing is obligatory.
It seems that what (1) expresses, an absence of obligations, is possible. For example, consider a time when no rational agents existed in the universe. Why should we think that any obligations—even those pertaining to tautological states of affairs—existed then? ... If we are reading OB as simply “it ought to be the case that”, it is not clear that there is anything counterintuitive about OB⊤ (now read as, essentially, “it ought to be that contradictions are false”), but there is also no longer any obvious connection to what is obligatory or permissible for that reading, or to what people ought to do.
That this implicit relevance-of-agents is understood as characteristic of SDL and its descendants, is indicated in the SEP article on Ernst Mally's deontic logic:
Mally was only interested in the deontic status of states of affairs; he paid no special attention to the deontic status of actions. Thus, his Deontik was a theory about Seinsollen (what ought to be the case) rather than Tunsollen (what ought to be done). Modern authors often regard the concept of Tunsollen as fundamental.
And yet for all that, McNamara and Van De Putte's article seems to characterize explications of agency modulo SDL as "revisions".
One might say that, historically, ethical (or where these are different, economic) theories grounded primarily in the concept of cardinal utility, would've been fleshed out by people beholden to things like positivism or behaviorism, or at least kinds of naturalism that yet adverted to classical logic, for the sake of a generically "realistic" mathematics (the basis for characterizing utility functions). Yet so John Rawls, for example, thought that appeals to utility functions intersected impersonalism while presupposing ethical factors illicitly (I'm having trouble finding where exactly, but in A Theory of Justice he says that utilitarianism presupposes moral standards for judging the very methods we use to represent utility functions "in the first place"); and Rawls frames ethics as "constructivist," ambiguously typed between mathematical and social constructivism.
But so then, besides a hazy, oversimplified historical account of why the more commonly used systems of deontic logic lack clear agency parameters as such, we can also appeal to (A) the commonality of classical logic and then (B) the commonality of SDL originally. What I mean is that if university curricula around the nation (America; I have hardly any clue what the prevalence of SDL might be in other nations) started including any introductory courses on some sort of deontic logic, they would probably have defaulted to whichever sort was easiest to teach. For better or worse, that happened to be SDL, and so SDL would've spread farther and farther, becoming more and more entrenched.
That situation could be compared to how calculus was originally taught in the style of infinitesimal analysis, evolved into the method of limits (with real numbers assimilated to Cauchy/Dedekind objects of the relevant type), but then branched off into nonstandard analysis with well-grounded infinitesimals resurfacing after about a hundred years; yet so that now, or at least for a while since Abraham Robinson's work, if you asked the average calculus student if they were using infinitesimals or limits, you'd've likely been told, "Limits," regardless of the limits(!) of those compared to infinitesimals.