Power cannot be stably balanced. But a power balance can be engineered into place. The latter is the easy part of sociology. Predicting what the attempt will actually accomplish seems roughly as hard as I expected it to be, but perhaps the antecedent soft problem can be transmuted. (I learned that just now.)
Moldbug's formally-specified sovereign corporation, with crypto-locks and everything, could be put into practice in the next couple months if just the right people decided to. (I suspect not that many.) That part's easy. Figuring out what it would change into later is the hard part.
It appears cultural organization is extremely flexible. If I put a car together wrong, it will jam or explode. If I put a program together wrong it will crash. Society doesn't seem to have a 'wrong' it can be put into. It will accept any sort of order as long as the order is being put into place by powerful enough entities.
Thing is, since humans instinctively know so much about sociology, I know more than I think I know. I can confidently predict that, had I the power, I could put into place a monarchy. I know what to do such that society reacts by monarchizing.
To exploit this skill, all I have to do is ask myself how society will react to being monarchized. Or: don't just work out why crypto-locks aren't cost effective. Work out what would accomplish technological loyalty. Don't stop at one option, find most of them. Relevant question: why doesn't the US Army stage a coup?
When I mention that the powerful don't have the interests of society in mind, I do so because thinking is work and therefore they'll avoid it unless there's something in it for them. That no society has ever been well organized doesn't guarantee it is impossible, it merely guarantees that nobody who had the power to do so had the drive to do so.