> It does not address the possibility that there are systematic methodological shortcomings in the polls.

I think 538 does take this into account in terms of the uncertainty, which plays into their probabilities. I know they certainly talk about this.