You're not really saying "yes", you're saying "I either visited this site or my coin came up heads.". Of course this only works if the other person believes that you flipped a coin beforehand, but in this case it's part of the data-collection procedure, so not much trust is necessary.

Presumably almost nobody visits the verboten sites, and so if you answer "yes", it's far more likely that you answered "yes" because your coin came up tails than because you actually visited it.

Or, in other words, the question is not "Did you visit this site," the question is "Did your coin come up tails or did you visit this site," which is a perfectly safe question to say "yes" to.

I don't think this approach applies in such situations, and wouldn't be a good idea if it did.

It may help with categories of sites, like "did you visit a news site today" or "did you watch porn".

Systems that don't deal well with edge cases aren't so great.

For people who don't really need to care about privacy, I guess that differential privacy is good enough. But there's a gotcha there, for people who ought to care, but are clueless. I'm reminded of that ex cop in Philadelphia, who believed Freenet's claims about plausible deniability.

Edit: A prudent standard for calling something "[foo] privacy" is arguably PGP.