Good question. I will do my best to answer how, as Iowahawk always calls “9% response rate” at the end when I discuss modern polling problems.
Please ping me too regarding Part II.
BTW, I have often suspected many of the things that you said, namely that internal, private campaign pollsters would have to be looking for the actual truth and therefore would be going about things entirely differently than the publicly published pollsters, who for the most part look to me like they are paid to try to influence an election rather than actually predict it. Basically, it seems to me that the private polls are analogous to astronomy and the public polls are analogous to astrology: the former is an a scientific attempt at explaining the universe and the latter is for consumption of the weak-minded, or in the best case, for entertainment purposes only.