Analysis is more complex than just what percentage of ballots remain. It was widely expected for Pennsylvania to have a decisive swap when Trump had a seemingly huge lead and Georgia a very red state was predicted as being a vastly closer race. North Carolina could have also been called as leaning Trump much sooner, but overall it was a fairly close race.
Voting by people in large groups are predictable and people spend a lot of effort getting this right. Out of 50 states exactly one was significantly different from predictions based on a significant shift in a specific group which is easy to miss in statewide polls.
> Although Biden's lead has only grown there saying that it's "obviously won" is a touch optimistic.
No, if you follow the election analysts like Nate Cohn, Nate Silver, Dave Wasserman, and especially Jon Ralston (who is "the guy" for Nevada elections), Biden has been the clear winner for nearly two days.
And also AP almost immediately called Nevada after the Pennsylvania drop, so they were definitely waiting on that.
And might have been right, Biden's lead in AZ keeps shrinking, 3 days ago his lead was 25% of the ballots left to count, now it's 15.
> And then Nevada has obviously been won for days, but nobody was calling it because that would have put Fox and AP over 270.
Although Biden's lead has only grown there saying that it's "obviously won" is a touch optimistic.