Boiling it down to its bare essentials, I think the question is really whether or not it’s plausible to claim that getting potentially unlimited resets, but still tracking each entity from a finite pool, with the ideal goal of any one given entity changing on each reset in a non-deterministic way — there would eventually come a state, in which each and every single entity has, at some point, encountered their then active ideal goal.

If we don’t track the entities and/or the pool is not finite, then I would say it’s simply impossible. But even then there are boundaries and variables that need further defining.

But if we assume the initial scenario I described, then sure, I think it is inevitable that a finite set of beings will eventually have all experienced their ideal goal, at some point, assuming the goals are, too, finite. And even then the one limiting boundary — time — would have to be infinite. If not, then we’d also have to define the entire thing further.