Pretty sure that 2020 polling was more or less spot on in Virginia (slightly overestimated Ds), while it slightly underestimated Ds in 2018. This is not WI or MI, where there's a consistent pattern of massive D polling biases, surely not to the extent that we'd need for a Youngkin win.
It actually didn’t underestimate Ds in 2018 (at least not according to the Atlas polling database, Wikipedia, and RCP). Either way, you clearly implied that the race being "within the margin of error" (not a fan of this characterization) or within 4-5 points didn’t matter because polling underestimated Democrats in 2017, the first VA election under Trump (with undecideds less likely to break Republican than under Obama and turnout being more difficult to model).
I agree with you that polling tends to be more accurate in VA than in the Midwest and other parts of the country, but then why dismiss the multiple polls showing the race within 5 points by referring to the 2017 "polling error"?
Mark Warner underperformed by a significant amount in the 2020 election. Gade did comparable to Harrison in South Carolina.
