As most rational /. readers must have noticed, the expression "correlation is not causation" is mostly used by ignorant apes looking for a quick +1. Science stories get tagged with it even when the story itself is only about the effect, and points out multiple causes, or in a story about the discovery of an indicator, as if it had implied bromophenol blue causes acidity. The expression seems to have been entirely taken over by people with no knowledge of science or logic, and it's usually used as a strawman argument against the science (and through that proxy against more evil things, often environmentalism or socialism (as biology, medicine and social sciences, etc., are considered socialist, as opposed to the libertarian idiocy so many Slashdotters believe in)).
The harmfulness of "correlation is not causation" is thus two-level: on the one hand, it drowns /. in illogical and retarded drivel that too easily gets modded "+5, insightful", on the other hand, it's usually put forth in defence of potential harmful practices in real life. Remember, there's only a correlation between smoking and an untimely death.
Oh, and what can we do about it? Personally, I pledge to use my modpoints by modding down whichever comment that first makes use of the meme in every science story, whenever I have modpoints. And you will know me by "-1, redundant" or "-1, off topic", whichever seems most applicable.