The most interesting part of the polling process to me was hearing that not only were they looking at the span of numbers 1-5 on individual questions and individual monsters... they were looking at the span of numbers across all monsters.
One of the repetitive complaints about Wyatt's polls were that they always were dominated by results in the 3-4 range. So just looking at those by themselves, it was always "Well, the polls were written so to get the results they were looking for! Self-fulfilling! Meaningless!". When in truth... the proof of the pudding was in comparing monster results against each other. So that while yes, most results still fell within 3-4... there was definite indications (like Wyatt pointed out) where 4s spiked. Or 3s spiked. Or 1s and 2s spiked, pushing the 3s and 4s further over. Thereby telling him (for instance) that the Good Monsters article was not up to stuff in comparison to all the other monsters he looked at, and thus some rethinking was in order.
Taken just on its own... a person would look at that Good Monsters article, see that the range was still primarily in the 3s and 4s, and thus think there was no worthwhile info gained from it because most votes were still in "Right track!" territory. When in truth, there was more info behind the numbers than we all were realizing.
We are people of hyperbole nowadays. Everything has to be "The worst thing I've ever seen!" or "The greatest thing ever!" As a result... unless these polls showed massive amounts of 1s or huge numbers of 5s... everything else was going to be read as just middle-of-the-road results that were to be expected, and thus nothing noted or gained of any worth. "Useless polls." Which also explains to a T so many of the arguments on boards like these... where you can't just offer moderate responses on either side of an issue... everything has to be EXTREME!!!! Because we think that only results that fall on either far end of the spectrum are "real". Anything in the middle is an 'Eh!' and thus too easy to ignore.
Whereas quite honestly... being able to read the tea leaves of the middle results actually can get you closer to the truth.