Your incessant use of this canard only weakens your arguments. bitcoinocracy is no more credible than consider.it.
Bitcoinocracy is immune to Sybil attack, because you must prove ownership of BTC to participate.
Consider.it is so open to Sybil attack it might as well be a honeypot.
The former is signal, the latter is noise.
You're just butthurt the deep pockets support continued consensus more than contentious hard forking.
If Bitcoinocracy was in favor of Gavinista governance coups and contentious hard forks, you'd never shut up about what the "community" wants/demands/deserves.
They are all volunteer samples, so really it will only get people who care very strongly one way or the other to actually respond. You aren't going to get the opinion of the average bitcoin user but rather of people strongly in favor or strongly against. It isn't representative of what the population really thinks and thus it isn't consensus
Those platforms are vulnerable to sybil attacks, though Bitcoinocracy less so. People can make multiple fake accounts and thus submit fake votes making one option more popular than it really is. Bitcoinocracy combats that by requiring that people sign messages with addresses so it really is based on how many coins. But that gives unequal weighting to the votes, although I suppose that is a better representation of the economic majority.
The questions asked in those polls are not without bias. The way that the questions are phrased typically imply one meaning or the other, kind of making people want to vote one way or the other. The questions should instead be neutral and bias free to get the most unbiased responses. Also, the questions should provide information for people to read about what those changes are actually for and how they will work so that people actually know what they are voting for. The information should of course be unbiased, but that information is not available in the polls.
In any scientific study, just having one of the above biases would invalidate the entire study. Therefore, since any online voting platform has those biases in them, the entire voting thing and checking those sites for the supposed consensus is invalid and thus should not be considered or referenced when discussing consensus among the population.