A tale as old as time. I don't just want to know the instances of harm, I want to know how representative they are supposed to be. And I think people who have thought through the latter question aren't the types to do the former.
In neither cases it was really a flaw of wikipedia, the model is clear, contributions should include content and sources, and it's the responsibility of the reader to check the reference.
I get it if bullshit is uploaded and a layman is fooled, but both cases involve trained professionals who know very damn well what a source is, judges and journalists, and it was principally their failure, not of Wikipedia.