If a mechanic doesn't help with making it a better game and goes against historical possibility, then there is no need for it.
That's not good criteria. A player who want to create trade company in Ireland by England definitely can say it would make game better (because good game is a game player get fun with), and he can definitely argue they were historically possible (Sweden can do it in Lappland, why couldn't England in Ireland?).
Argument "Trade companies are supposed to represent entities that engage in very profitable large-scale and long-distance trade" is good, but it isn't about "is game better with it" or "was it historically possible", it's about simulation integrity.
To bring all this philosophy closer to topic - we have a question about changing tags.
Why limit tag switching in the beginning? Players switching tags gain a lot of permaclaims around. If you ask "is it making game better" - it's subjective; if every player believed its not tag switching wouldn't exist. If you ask "was it historically possible", it would be long, senseless discussion I love. But if you ask "what it supposed to represent?", everything became clearer. In my option tag represent national identity (that's why we have national ideas), and therefore permoclaims should be changed into special wargoals some tags get, like "manifest destiny" cb in CK2.