Mechanically, issues are just a colorful word for “game session” and after a number of sessions (usually six) you “level up” by gaining a collection, which gives you a once per session benefit when you choose to use it. Usually your GM will have at least a loosely-planned story in mind for an issue, with most of the planning being a possible sequence of scenes and most of the prep being somewhere from one to three possible action scenes. IME most issues only have time for two action scenes, sometimes only one, but that can vary a lot- especially if your game sessions run to twelve hour marathons or something rather than the more common three or four.
Playing “sandbox” the way you’re describing would really just be a matter of the GM (with some input from the players about what they and their characters are interested in) coming up with starting points for several story ideas rather than one and letting the PCs learn about them diegetically, then decide what one(s) to pursue. Beyond that it’s really going to come down to a developing narrative, so you need to be flexible and think on your feet because your players will surprise you, guaranteed.
There’s lots of things you could do as a GM to push things forward overall. You should probably set up some sort of loose campaign calendar (“publishing schedule” would be a decent term) that advances as time passes in game - but different stories might take different amounts of time. Responding to a warehouse robbery that leads to the villains being captured is probably very little calendar time. One that results in the villains escaping, a scouting montage/investigation scene where the players track them to their lair and raids the place, only to find most of the baddies have left for another heist that the players could either race to or wait and ambush them when they return - that could be a few days or more.
Maybe some storylines are on timers and will advance (along with their villain’s plans) as calendar time passes even if the PCs are following other stories, eventually culminating in some kind of crisis that intrudes on the players’ plans whether they like it or not. Same goes for preplanned events like a natural disaster or invasion of mole men or whatever.
Certainly takes more work than a more linear play style where there’s really only one storyline on offer at a time, but you can’t be accused of railroading the PCs. If they ignore increasingly urgent tips or requests about a dangerous plotline that’s progressing while they do something else they’ll find it catching up with them eventually. And of course they’ll probably unintentionally provide you with all sorts of ideas for new stories to start up as they do stuff during the original options.
That’s about all I’ve got offhand. I usually take a compromise approach by plotting out at least two sets of things that could happen in a session and hoping the players will jump on one of them, often with an “aggressive” plot that’s coming for them if they don’t want to engage with the other options. For ex, they might have a mystery from a previous session they could look into, or they could go help a police contact with rumors about a new villain in town, but if neither one draws them in there’s some villain with a grudge looking to get revenge and he doesn’t care if they wanted to stick around their HQ training in the Doom Room. Maybe the group even splits up and deals with all of them at once, in which case the plots may run into next issue or even tangle with each other.
