DnD has had zones of control (ZOC) and the attendant attacks of opportunity for a very long time now. It was ill defined in 2nd ed, far too lethal in 3rd ed, but got a lot better in 4th and 5th. The biggest issue with ZOC is that they create an arms race between melee and ranged characters that spirals out of control and becomes far too defining. Instead of making it a deliberate choice whether or not to try to pin someone down the combat game often comes down to figuring out whether the melee character has more ways to keep in close than the ranged character has to get away and the result of the combat turning on that single calculation.
Naked Man recently came up with a bunch of ideas for making ZOC more important in Heroes By Trade. He really likes the idea that once you engage someone in melee they can't just walk away trivially and he wants there to be a cost for retreat. The idea that two combatants are really in the middle of a swashbuckling contest and not just trading blows sequentially does feel good and making retreat challenging supports that. His solutions are very much like DnD in that people get to make attacks of opportunity when someone moves away from them but the package includes ways to get around that by skipping your Action for the turn.
There are real issues with this sort of implementation. For example, it lets a single large creature rush in and put their ZOC on an entire group, effectively pinning the entire combat in place. Nobody can afford to take a full extra attack from a single enemy nor can they afford to waste their turn walking away so they just stand there. This isn't improving realism any because it isn't as though that ogre is meaningfully engaged in melee with six enemies at once, much less able to punish them all for retreating. You can deal with this by restricting attacks of opportunity in some way but then you have to deal with facing, recording attacks of opportunity, or finding some other complex solution. I strive very much to avoid complexity of mechanics particularly when it is overseeing trivial tactical decisions.
I think that complexity might actually be the real issue here. If I design ZOC so that people can just ignore them because the penalty is low then all we end up doing is a lot of math for no tactical impact. If I design them so that ZOC are critical then people will just stand there and brawl to the death as soon as they get adjacent to someone and that isn't the crazy mobile combat I want to achieve. I don't know that I can actually hit the middle ground without ramping up the complexity so much that it isn't worth it. You only get so much complexity before people just give up and tune out and I don't think this is the spot for it.
However, all this does make me question my current ZOC implementation. Right now ranged characters take disadvantage on Hit Rolls while inside a melee character's ZOC. That reduces their damage by roughly 35%. It feels okay to me because it seems thematically sound that aiming a bow or a magical blast is harder while desperately dodging sword swings. Of course you could make the same argument for dodging arrows or fireballs too and I don't do that at the moment. This solution also does not restrict melee characters from wandering about at will.
I am hesitant to remove the system entirely though as if ranged characters don't have any sort of disadvantage when pinned in close I would have to amp up the numbers on melee characters massively to compensate. I don't so much love the idea of melee characters getting whittled down as they rush in and then instantly annihilating any ranged character that failed at kiting. That sort of thing just seems way too swingy.
My main concern in adding something that will consistently force melee characters to engage and then just stand there fighting to the death is that melee have traditionally been stuck with uninteresting choices in fantasy games and I really want to get away from that. I like the idea of them dashing about and being able to decide who to bash on.
Taking all this into account I think the current system is the best one I have found. It puts some pressure on ranged characters to make up for the simple fact of having ranged attacks and it allows melee characters to make lots of choices and be mobile instead of being pinned to the first thing that swings at them. Those two things seem like things I want and achieving those goals with a fairly simple system feels good.