I think at some point it becomes necessary to back away from the metaphors, when they begin to cloud the issue. So for a few paragraphs, let's try stepping back and looking at the thing itself.
With my last post, linked above, I reached the distinction that most games compel the players to be reactive instead of proactive. I'm quite sure that most of my readers know precisely what I mean by these two terms, but let's define them anyway, because with a good methodology, we have the best chance of producing the best results. Think of it as responsibly dotting our 'i's.
To be "reactive" is to respond to a stimulus. And so it goes in most games, and in most game advice I've found. The argument is made again and again, from Matthew Colville to Matthew Mercer (should I change my name to 'Matthew'?). Give the players something that will stimulate them and have them react.
To be "proactive" is not merely the opposite of reactive. It is to create or control a situation by causing something to happen rather than responding to it after it has happened. The underlined is the key in this case.
Consider. Reactive brings to mind words like recoil, talk back, echo, backlash, imitate ... reflexive words, suggesting a lack of thought before answering. Proactive suggests words like enterprising, energetic, driven, bold, dynamic, motivated ... strong, heroic words that we like to imagine coming out of ourselves.
Yet all the advice is aimed at pushing the players to react. Why?
Well, to begin with, it's easier. A stimulus, a thing or event that evokes a specific functional reaction ~ and, for D&D, an active, energetic reaction ~ requires little effort. The door busts in and a bull starts attacking the people in the tavern. Someone is collecting dead bodies and threatening to invoke a god. React! Do something! GO!
Inspiring someone to be proactive is a much, much harder proposition. You can't just light a fire under their ass; you've got to approach the whole issue with an argument that causes a light to go on in the player's head, causing them to get an idea. The idea, in short, can't be yours. It has to be theirs.
If we present an adventure to the players, either bought or self-produced, we can minimize how much data the players need, enabling them to react. But if we want the players to be self-energized, we have no idea what information to provide! Because we have no idea what will produce an insight on their part.
So we have to produce so much more pure and direct data than we would ever have to invent for an adventure ... and we know from experience ~ from the vast number of people condemning the agenda of this blog, for example ~ that mountains of pure and direct data is just too hard, too time-consuming, too confusing and too much work to produce. Easier to make the players react and get the damn show on the road.
Sorry, that was a metaphor. It got away from me.
Before we can hope for a "sandbox" campaign, however we want to define that metaphor, we've got to have a sandbox. And if you sit down on the edge of the sandbox, all of the sand is right there, ready to be used. The procedure, the tools you'll use to remake the sand, everything you need, is ready for you. All you have to do is look at the sand and decide what you want to do with it.
Let's take a game world, then. And let's say we're sitting at a tavern, because it is a massively overused trope (that I never hesitate to use). Forget the DM, forget your character's backstory, forget everything about what your character can do at the moment or anything about how your characters happen to know each other. The only thing we want to consider right now is this:
What do you know?
If we're sitting in most worlds, almost nothing. We might have a vague knowledge of being in the middle of a town, surrounded by some sort of wilderness. If this is the Keep on the Borderlands, we know that the forces of chaos are pressing upon the realm's borders, that adventurers find their way to the Keep in search of adventure, that we're stout-hearted, that there are dark forests and fens, that there is a place called the Caves of Chaos, and that there are a whole lot of civilians and guards who seem involved with whatever it is the Keep does, but who are probably not going to help us.
That isn't much. If we sit four players at our game table and expect them to "come up with an idea of what to do" on their own, based on this inconsiderable information, it isn't hard to guess what that's going to be. It isn't enough to just say to the players, "You're here. What do you want to do?" You've got to make the world itself meaningful and intricate enough that there is at least a chance that they'll put a pile of bits together in their heads, add water, and feel motivated to make something happen.
But we don't do that, do we? We spit out a few obvious "ideas," make claims that game worlds are "amusement parks," and then wait to find out what ride the players want to get on. And if we go so far as to offer them three different possible rides, we call it a "sandbox" and pat ourselves on the back.
Look around, right now, where you are reading this. If you close your computer and stand up, how many things can you think of to do in the next half an hour? How many things can you think of that don't require asking someone else to perform a service for you, like selling you coffee or fixing your car?
The answer should be lots, particularly if you're an avid DM. You probably have 15 or 20 unfulfilled plans of your own making, the starting of which can be done on your own, because you've invented the idea out of your own head. Now, why isn't your world designed to let the players do that?
It isn't enough to dump a few toys in a box. The box needs sand. Lots and lots of sand. Not just the DM's sand, either, but sand that comes out of the player's mind, because that sand has to exist too. If the player argues that there must be something reasonable in this town, in this region or in this world, the existence of which is entirely intuitive, then that too is part of the sand that makes up this box. Not because the DM condescends to say "yes," but because it must be so.
Only then is there a chance that the players will really be gaming.