[Goladus]

Contrarian anecdotes can be very useful (and logically sound) when they are contrary to a generalization. If someone asserts "f(x) = 5 for all x in A", "f(a) = 3 and a in A" is a logical contradiction proving the assertion false.

And, frankly, if we're talking about public discussion of scientific papers, inappropriate generalization is as big a problem as contrarian anecdotes, if not bigger. Scientific papers often cover highly specific observations that are useful primarily to other researchers, and often even then not for many years. People then try to apply that specific knowledge to practical day to day situations.

[Symmetry]

Have you ever seen that sort of generalization in an article on Hacker News? Even when you have situations like, e.g. nobody has ever recorded a case of a person with a certain genotype being infected by a norovirus, you don't see papers claiming that its impossible.

[Goladus]

Yes. Realize that often the generalizations are a lot more subtle than your example, and of course anecdotes aren't always the best or only counter-arguments you'll see, even if they are valid. Start looking and I am sure you will find examples.