I went through the survey yesterday and I thought the site's methodology was rather questionable. By showing the "most popular so far" responses first, it obviously drives more votes to those responses, while responses which aren't in the Top N (For, I think, N = 8? I didn't pay that close of attention to the chunk size.) aren't displayed at all unless you explicitly request to "see more".
With results that will be so clearly biased by the survey structure itself (even without considering the content of the responses), it really only seems suitable for "fun", "random" polls, not for any kind of serious data-gathering which might be used to influence the policy or direction of worldwide communities or an organization which (somewhat controversially) attempts to guide them.
