I was surprised, a few days ago, to see a post from ZapperZ recommending a Wall Street Journal article on quantum entanglement. It was surprising not only because it's weird to see anything in the WSJ that doesn't have an immediate financial connection, but more than that, I was surprised because the article contains a lot of statements that are the sort of thing ZapperZ usually denounces as unforgivable ignorance by journalists who shouldn't be allowed to write about science.
Happily, by sitting on the article for a couple of days, Tom said most of what I would've said, and I was saved having to write out a long explanation. The basic problem is the same thing that plagues most pop-level discussions of entanglement: the author talks about it as if the states of the entangled particles had definite values all along, when in fact the states are indeterminate until the instant that one of them is measured.
Their choice of words also skirts along the edge of suggesting that changes to one particle after the initial measurement will lead to changes in the state of the other particle. This is manifestly not true, and I would've liked them to be more careful about the description.
Do I think this is evidence of unforgivable ignorance on the part of journalists who shouldn't be allowed to write about science? Not really, no. It's a pretty good attempt, as such things go, and you can find much worse descriptions of the phenomenon without much effort. It's still a flawed explanation, though, and flawed in ways that would keep me from praising it too highly.