These sound like interesting games, but they make no mention of one key component I'm after: play acting the character.

Why should they? That's not the job of the mechanics. It's not the job of the system. It's not even the job of the platform.

It's your job.

Your job is to play the character as you can see if it. Your job is to come up with that character, not have it handed to you. Especially in the context of a simulation, the responsibility for characterization is on you. It can't be anywhere else.

I checked ME forums, and the Roleplaying forum is a ghost town. I suppose I could create my own server and go around trying to recruit people to play on it.

If you want the kind of role-play you seem to demand – that's your choice. Find a group, put it together, find people interested.

I'm trying to figure out how anyone could provide a game that would satisfy you, here. You don't want to create the character but you want to create the character. You want to drop into a setting that already exists, but you don't like the settings that exist. You want architectural freedom, but you don't want others to have the advantages of that architectural freedom. And you want things to be dangerous, but you don't want to actually be in danger.

Stop me if you've heard this one.

If "getting better at escaping danger" isn't a skill a player can develop as they play (as it is with all decent RTS and FPS games), then the game is doomed to fail.

And, oh yes, somehow having a labyrinthine abstract interface which doesn't actually represent the environment and doesn't give someone enough information to make decisions about which way would be more or less advantageous, and thus give them feedback in order to become "getting better at escaping danger" is superior to literally immersing the player in the first person perspective or third person perspective of their avatar representation and basing their "skill at getting better at escaping danger" to learning the actual environment and how to use it.

Are you sure you actually want to play a game?

Are you sure you actually thought about what you want?

Because from the point of view of a designer and someone who has no small experience of the industry, you have been repeatedly contradictory about stating what you want and what gratifies you in play.

If I didn't know better, I would be forced to observe that you don't actually want to play a new game. You want to play a game that you've already been playing, and which no longer really exists because other people have been playing it, and you have no interest in actually learning or immersing yourself in a new thing. You want the same old thing, despite the fact that that same old thing isn't really what anyone else wants, and rather than learn the new thing and what it's good at, you'd prefer not to because it might sully your memories (inconsistent and self-contradictory as they are) of the old thing.

Not to put too fine a point on it, but you are the kind of player that designers deliberately avoid catering to. Nothing will ever be good enough. Nothing will ever provide you the pleasure you want. That's because the experiences that you had are the best ones that you ever want to have.

The Darksun setting of D&D is my kind of game. It has been ever since it first came out. If I could somehow create a multiplayer version of that, with players actually play acting the different factions, I'd be in heaven.

As we say in the business, "see to it, fancy lad!"

There are plenty of open world survival games. Some of them are in fantasy settings. Very few of them are in high magic fantasy settings, because anytime you introduce high magic – intelligent people start to wonder, "why is it that you have this problem which is easily solved by this magical effect on a daily basis?"

This is a problem for you, because of all the things that you've expressed that you want, a consistent setting is going to be the hardest thing to acquire. Especially in the context of "a simulation."

You have a choice: detail which leverages the day-to-day activity of existing in the setting, which is going to require a ridiculous amount of work by the developers to just cover the details of mundane existence – adding on significant magic is a task almost no one is up to and certainly no developer is up to doing so on a purely systemic mechanical basis, or you can focus on narrative verisimilitude – but that would require you to give up all of the "simulationist" mechanics because they don't lead to narrative verisimilitude. They lead to mechanical verisimilitude.

So here's what you need to do – first find people. Find people to play with. When you find those people, find out what they're interested in playing. Except that they are probably not going to be interested in the same things that you are and you're going to have to make compromises. And then play something that exists.

Alternately, you can just not play.

Those are the choices on the table. That's what you've got.