Problem with surveys

I've heard it suggested that surveys are biased towards the more extreme opinions, that they don't take into account the opinions of the apathetic.  A recent survey taken across the majority of the United States has shown that, out of a test group of around 1100 people, 100% of them performed surveys.  This would seem to show, beyond a doubt, that surveys represent the entire spectrum of subjects.  As such, any error during a survey would be solely caused by the person performing said survey and not by the nature of surveys in general.

Do a series of surveys, and as the sample size increases, the mean of the distribution of the results will tend towards the true mean.