1. I get that... it's strange though... you could conceivably use it to get a random sample if, say, you polled people from Fredricton via their profile also. Just a simple mind shift.
2. I leave everything. I feel like privacy is exploited by those in power sometimes. Like salaries...
Oh, I see what you're saying in the first point now. These Facebook samples aren't exactly random. That's why I was mentioning the selection bias thing (do the missing people have characteristics in common). However, they're pretty great at getting a targeted sample.