Anyway Data for Progress is back on their accuracy streak...though now with exceptions. Michigan, Idaho, and Missouri were all close to the final result. In this environment we should probably expect the final Washington results to be near their Biden+6. However, the reservation I alluded to was that their have been less accurate in Black Belt states with rural AAs, and the Mississippi poll here behaves similar to that trend.
Yes. It seems clear that they have a particularly difficult time polling states with high percentages of Black populations (especially rural ones).