When communication was expensive you could only talk to people in your physical vicinity. If they had stupid ideas, you had stupid ideas. A town even a few dozen miles away would have an entirely different set of stupid ideas.
What's different now is that the clusters aren't based on geography, so there is more overlap and you encounter people whose set of stupid ideas are different than yours. That's actually a good thing -- it's how a lot of them get stamped out because they're so obviously bad that they can only survive when no one is even attempting to challenge them.
The problem now is, instead of there being a third world village full of people who think crazy things and whose children die of preventable causes but nobody else has reason to care that they exist, now there's a few of those people in every town getting infected with measles and then showing up to school with your kid. So now we have to care more about idiots being wrong because said idiots live closer to home than they did before.
But the way to fix wrongness isn't censorship. That's how you make wrongness increase, because as soon as you erroneously censor something true, nobody can say it in order to argue that it shouldn't be censored.
The way to fix wrongness is education and debate. Which is real and continuous work instead of a switch you can flip and make the problem go away. Sorry.
Using the postal reference, something may have taken a week to transmit to 100 people. Now, it takes 5 seconds.
That's a fundamentally different situation, and it requires different tradeoffs in society.
Is this even true? There are a billion people on Facebook, but it's not like a given post ever gets seen by more than a small fraction of them. Nearly the entire human race has been reachable by postal mail for over a hundred years; the size of the network really has nothing to do with anything.
What has changed isn't the size of the network, it's the culture. The CDC and a thousand independent doctors say that vaccines don't cause autism, but too many people don't believe them over the mother of a kid who both has autism and was vaccinated, even if that mother has no real evidence that one caused the other.
The problem there isn't that someone is saying something which is wrong, it's that other people are too credulous of unverified claims. Nobody should believe outrageous claims from some random stranger with no domain knowledge and who is contradicted by multiple domain experts, unless they're stricken with enough curiosity to actually replicate the experiment themselves. And even then they shouldn't believe based on the word of the stranger but rather their own observations or those of someone they themselves find trustworthy.
The solution isn't censorship, it's to teach people how to properly evaluate evidence instead of being credulous and easily manipulated fools.