Ann Selzer is an A+ pollster.
I take it with a grain of salt, but if it plays out it would arguably show the other pollsters 1) overcorrected for prior errors (which is traditional and explains why polls never miss the same direction 3x in a row) and 2) are unwilling to be outliers and thus adjust their pools to match others, creating a self-aggregating herd that will all miss the same way.
Anecdotally, we took my daughter to a roller rink for a birthday party in small town Iowa today, and while the Trump signs were bigger, we actually saw more Harris signs. A friend was in Dubuque, which had one of the biggest Obama->Trump swings in the country, and they similarly saw more Harris signs there than Trump ones.
Do I think she ultimately wins Iowa? No. Could this mean the Dems pick up some state leg seats and hopefully 2 of the 4 Congressional seats? Sure. We could know early on whether the Des Moines based seat will flip. Polk County is normally very quick with their absentee votes reporting. The Dem needs to win that county by at least 40k to stand a chance at carrying the district, because every other county will go red. Absentee votes will need to be at least 30k in his favor.