you need to haul your ass back to school
apparently your dumb ass can't understand the difference between a statistical sample and a group of people who are NOT selected for statistical purposes.
here, let me explain: if you take 1000 british people that live in london and ask them where they live, it is NOT correct to say based on that that 100% of british people live in london. get it ? nah, i doubt you do.
those 34k people got together for a reason: to quit facebook. so even if all of them did, it doesn't mean all the users of facebook quit the site. on the other hand the percentage of people who expressed interest in organizing this protest might be somewhat relevant - just 34k out of hundreds of millions is a very small percentage, and even if there were 10 times more, it still would be next to nothing.
now if you want to have something resembling real science you can take 1000 RANDOM users of facebook and ask them if they intend to quit because of the privacy concerns. of course it wouldn't be real science, cause your sample would influence the result, but this is what you have formulas for. formulas for deducting the proper sample size and the error margins for a sample. and of course proper sample sizes depend on the sampling technique, from random (basic, more error prone) to more advanced and less error prone techniques.
PS: what the fuck is this facebook i keep hearing about ?