to the question that we discussed – report makes no attempt to separate R- and D- leaning polls. They took all public polls on realclearpolitics, 538, etc and lumped them together.
They compare accuracy by type of polling: joint online/phone polling is only 1% off, while pure online and pure live polls are 5% off. 50% of these joint polls are by one company that they do not identify, so it seems that this one company has good methodology. I looked up realclearpolitics:
final general results are Biden +4.5, and closests polls are IBD/TIPP, The Hill, Emerson, sounds like one of them is that great poll. Rasmussen biased 3.5% towards Trump, the other TEN polls are 2.5 to 6.5% towards Biden.
So, no conclusion can be made about different polls in this report, except the above, because the report does not address the question. The expert who was quoted did not have any data to support in this report.
Report underestimates the level of the problem – polls showed 11% preference for Biden in October and then paddled down. So, I think, reasonable Dem voters should understand frustration of the R- voters at these polls (together with other media influences) that undoubtfully affected voter enthusiasm – given how narrow the election was (50K votes in Presidential, 50-50 in the Senate and the House)