@Amphebianlesbian regarding your extensive link-dump. I've cleaned it up by adding HTML so that it's easier to read.

Whenever I read an article that quotes a "Twitter used named X", I assume that the author has been trolled. Like the "Marie Christmas" situation i described in "The trouble with Twitter - Trolls...", it's just too easy for a random user to get the attention of a journalist and then never prove who they are. Parody accounts pretending to be X are a dime a dozen, with Godfrey Elfwick being a great example of a random Twitter user who keeps getting embedded into articles as if s/he is serious. A person on a flight can misrepresent a situation - "X is being taken off flight because Y (muslim/black/breastfeeding take your pick)" - with a poor image, and suddenly a global brand has to address hundreds of news articles to fight an accusation. One could start a protection racket with the right twitter accounts, "Nice brand you have there, would be a shame if I start trashing it on social media...."

With the case of Marie Christmas, at least some journalists were going the correct route and asking for verification of who the person was. Others dove deep into the timeline and saw it as "trolly". The problem is that journalists deadlines are now "two hours ago" by default, and this step is often skipped because other outlets are going ahead. That is how AP quoted a random twitter troll as if they had been an eyewitness to the San Bernardino attack.