We're talking about setting rules that span literally _billions_ of conversations. Taking fully for granted the notion that there are facts that are objective, 100% knowable facts: The question isn't about how the system handles obviously false or obviously true statements, it's how it handles everything on the margin. There were tons of things that were considered conspiracy theories that we now consider clear fact: people like you would have been the ones saying "of course we shouldn't allow people to claim the CIA is drugging and torturing and raping children"[1] in America in the 1950s, or "of course we shouldn't allow people to claim the government is euthanizing the disabled"[2] in 1930s Germany or any number of "obviously false" things that were very much true.
I have no reason to believe your particular example falls into that category, but the sweeping claim that arbitrating political facts is trivial is profoundly ignorant of even a tiny bit of history.
[1] Project MKUltra
[2] Aktion 44
Facebook is arguing for the right to poison the well with a literally unlimited volume of toxic disinformation, rather than figure out a way to reduce if not eliminate that harm. Perhaps they limit the number of political ads that any single account can view to some small number. Perhaps they can only fact-check ads on topics that reach a certain view-count level. Or if it's too difficult of an engineering problem for Facebook to not poison the well, then they can stop hosting political ads.