Okay, I'll grant you that. It's better to have a sufficiently strict filter that loses some useful information than a weaker filter which lets in garbage data. I would presume (or, at least, advise) that you make a particular effort to analyze data which you previously rejected but which remains widely discussed, however - an example from my own experience being Searle's Chinese Room argument. Such items should be uncommon enough.
Agreed.
Tyler Cowen argues in a TED talk (~15 min) that stories pervade our mental lives. He thinks they are a major source of cognitive biases and, on the margin, we should be more suspicious of them - especially simple stories. Here's an interesting quote about the meta-level: