I suspect the vast majority of these are nitpicks, so I wouldn't assume that they overturn results and papers. They don't fall into the normal corrections process, making it hard to understand how to handle them well within the constraints of academic publishing.