If the point of social ostracism is that it's good because it's old, then verifying whether this claim is true by comparing it to other old stuff is nothing shocking.
You'd need to make a better point for ostracism to be good, other than it being old.
Several problems here.
The first is, as mentioned by me upthread, that it is one thing doing this with people you know, as, in your example, Mary and Stan. Since you know them, you'll be able to assess every factor, including the reliability of Mary and the likelihood of Stan doing that thing. But here we're not talking about people you directly know. We're not talking about villagers in a small village, where everyone knows everybody and secrets are difficult to keep. We're talking about you taking decisions on the "word" of someone you don't know, about someone you don't know.
Secondly, imagine indeed that it is Stan who is your friend since childhood, and Mary his wife you don't know all that well. Wouldn't you talk with Stan before terminating your friendship? I would.
Also, I find it surprising that at this time in the history of the Internet, in which everyone is well aware of the dangers of fake news, there is no thought spent, with regard to this practice, as to the possibility that these phenomena get manipulated. Nor is this a "new" development. I'm sure you'll dispense me from providing the historical examples.
And finally, you're winning easily by choosing actions and behaviors we can all easily agree are deplorable. The thing with social ostracism is that it's driven by the majority - or, even, by the loudest minority, deciding what is right and what is wrong. And neither are always right.
