It seems to me people who are able to be cold called (haven't opted out/been careful with their phone number/screen their calls) will be much less privacy conscious than a true random sample
The Onion should really do a parody of studies like this. "100% of Americans Answer the Phone When Strangers Call" or something like that. As it is, dubious research like this just generates misleading headlines that then misinforms public debate about important issues.
Having said that, based on personal experience I don't doubt the basic results in terms of being for/against privacy, but the numbers and methodology are probably garbage.
I'd be very interested in any study that used different communication mediums to ask equal numbers of randomly chosen people the same questions. Phone, email, Facebook, etc. I think it would raise a lot more questions than answers.