I would like to suggest that the burden on building worlds should go to those who ask the question, rather than those who answer it. I understand Worldbuilding SE to be a place where you can come with a world you already have in mind, to ensure that it works how you think it does, or to see if there is a way to achieve what you want with the resources you have available.
In this case, I don't think enough of a world was provided. We get some explanation of what demons are, but nothing that could specifically be used in the context of the question. In essence, the question could be about how to control regular people by knowing their names. The problem that arises is that when you start adding in supernatural elements, answerers start to make assumptions about your world, and they start building their own interpretation of what they think you want.
What I think this leads to is a list of different worlds, where (in this case) a whole host of different kinds of demons are controlled via a whole lot of different methods, all having something to do with names. This isn't helping someone build their own world, it's giving them a bunch of pre-formed worlds and letting them pick the one they think is the most interesting.
This is why we ask for criteria. You need to flesh out the world you want, the systems you have available, the fundamental laws of your universe from which answers can be built. We may build worlds here, but we don't build them for other people, just like Stack Overflow doesn't do people's homework or write their software for them.
Now, this is just my understanding of the issue; there have been a lot of highly popular questions where the asker has done almost no work of their own, where every answer is its own world, and where there's really no good way to pick the best answer. If we want to destroy our integrity just to allow these questions to keep on being asked, then I guess I'd be okay with that. However, I do not believe that's what we're here for. I think the case studies proved that there are a lot of questions that can be asked that are interesting and useful and definitely on-topic and suitably constrained. But who knows, maybe I'm wrong.