Oh, they care once they really understand. The problem is, most people, aren't really thinking that critically about it until it smacks them in the face at which point its (usually) too late.
At the end of the day, John Oliver hosts a comedy show. Though it tends to be on the informative side, it still oversimplifies and exaggerates to get its effect. The more I've watched him, the more I've grown tired of his explicit style of "share a factoid before immediately following on with a forced pop culture reference" that summarizes his work.
He seems to be becoming more popular as a research substitute, though.
I suppose but I've seen literally the exact same reaction when I've explained things like this to people in those sorts of terms vs. the terms the news uses.
The problem is the "official research" doesn't explicitly state things like that but use it in terms vague enough people really don't believe/understand that the pics they send their lovers are being stored.
> http://www.gallup.com/poll/163043/americans-disapprove-gover...
Posing the question like that has serious biases because people really and genuinely do not understand the implications.
We need official research that is clear, concise, and written in a manner where the connections are obvious. Official research that relies on statements/questions that are vague enough to be heavily biased by the sources of the media the person consumes is valueless, honestly.
edit: also, are you really surprised that a comedian is better at explaining important issues to ordinary people in a way they understand than a security researcher or academic? He wasn't attempting to take a sample, rather describe how the issue should be framed. I wouldn't take it as evidence of anything, but rather a heuristic for arguments with your less-privacy-conscious peers, and it's a pretty solid heuristic I think.