Don't forget the Dunning-Kruger effect - sometimes the reader scoffs because they reversed cause and effect, or saw a correlation that doesn't exist ("vaccines cause autism!")

Additionally, quality on one subject by one author says little about overall quality. Maybe the author was forced to fill in for a sick subject matter expert, or rush re-writing the article after a hard drive failure, or just don't care because layoffs are coming... maybe the paper poorly interviewed their sports writer because sports is a filler for that paper. Maybe lots of things.

If you're trying to avoid Gell-Mann Amnesia, you need to trends of quality not single data-points.