1. Well, you didn't know that the rallies were indicative of enthusiasm. I didn't either. I thought it was a lot of bullshit. I'm not sure it's not bullshit, but in any event, I don't think I was ignorant in 2016. I think the knowledge didn't exist. I don't think you were ignorant.
2. This whole idea that response rates are down and therefore the polls are bad -- I don't know if the conclusion follows from the premise. Low response rates mean that pollsters have to work harder to get their samples, that's true. But it doesn't necessarily follow that the samples they get are somehow systemically biased. With all the polling we have, only systematic bias is likely to get through a poll average.
Could the low response rates create systematic poll biases? Sure, it could. I can think of a lot of plausible stories consistent with that idea. Maybe Trump supporters have become more eager to talk to pollsters because Trump talks about the polls so much and they want to push him into the lead. Maybe the people who used to respond to polls have had enough, and thus bias in the polls in previous elections could therefore cause the polls to bias the other way (though that is unlikely because of numbers). It could be that the total universe of people who answer polls has shrunk so much that the pollsters are basically just sampling the same people over and over again (though this story also potentially runs into number problems).
But we don't know any of that to be true. It's also possible that the samples pollsters have obtained are, in fact, representative and the only issue was the difficulty and cost of obtaining them. The fact that the polls are commonly weighting by recalled vote, and the polls that don't tend to favor Kamala in the EC, gives me some hope. That's an unreliable method and I think it's being used as an ad hoc patch of uncertain validity. But again, we don't know.
