You need to learn to read more carefully and with understanding.
You are trying to make methodology sound scary with some convoluted blather. Of course poll models are weighted for demographics. There are not some random "separately conceived ideas", they are based on statistics, past voting patterns, census, etc. There is nothing wrong with that. It's the science behind accurate polling. It's precisely the reason you don't need huge samples to get accurate results. It's why polling has gotten so good and your theory that it doesn't work is silly and wrong.
Stop trying to change the subject. This is your argument:
As I've been explaining since last summer, the use of polling to determine public opinion has collapsed to utter nonsense.
November 2012 happened long after you apparently began making these silly claims. The polling results for the 2012 general election cycle destroy your theory completely. The polls were accurate. Your theory that they aren't was entirely incorrect. You're not going to be able to cover that up with a bunch of psychobabble. You obviously don't really understand polling and just don't want to admit it.