supplementscene: But you can't deny it isn't an advantage for a fascist to nominate his Hitler as chancellor
For the sake of argument, assume this is right. Note that getting FFF cards, which isn't improbable, means a fascist and Hitler get put in bad light with one stone.
But, as I said for the sake of argument, say
it is advantageous for fascist who is first president to pick Hitler as Chancellor.
Here is the thing: these kind of games generally have a good negative feedback loop, which "smooth" out such advantageous plays.
Taking the example of our assumption, if a fascist takes Hitler as his Chancellor, and is later revealed as Fascist, then his chancellor comes under scrutiny and if is indeed Hitler is put under bad light? Why? Because
it is advantageous for fascist who is first president to pick Hitler as Chancellor The problem with you is that you see only the first loop iteration: if we taken the choice away from Fascist, he won't choose Hitler, but you refuse to see the feedback loop.
If we take the choice away from Fascist, then if Hitler is chosen we won't be able to analyse his choice and thus figure out who Hitler is. If Hitler is in the right spot, he's chose with a 100% certainty and our weapon has been taken away from us.
In general, if X is useful for fascist/mafia, then they can try and do X, but doing X reveals them as fascist/mafia, so they should be careful and not do it... etc. The more advantageous it is, the more careful they should be not to do it too often. Negative feedback loop control at its finest.
All that aside, in this game people can simply vote for the government they want, so if the majority want "meta" they can vote for "meta". I don't think that's the case here. I know you're obstinate in such cases but maybe we can just agree that regardless of whether it's good or not, "meta" is not happening this game?
Personally, I am very open to plays that skip governments and cycle back to a given player, but if one wants me to go with them, one has to convince me properly and back arguments with some facts. The whole case for "meta" seems to be built upon logical fallacies and statistically dubious arguments. So, no.