Let's assume the situation were reversed. Do you think the pundits would stick to the "No, fraud is not widespread and there is no evidence" mantra? Or, do you think the anomalies presented would be scrutinized with the precision of an electron microscope?

I was watching primarily Fox News during the period of time in question in this article, and Arnon Mishkin and others from their decision desk explained all of this. They expected mail-in ballots to be overwhelmingly Democratic in all counties in all states, which they were, because Trump had said not to vote by mail-in ballot because it was insecure and subject to fraud. The article author mentions none of this as a factor for the skew.

Democratic dominate precincts are also significantly more populated, so you get much larger tranches. There's a variable stop in the count because some counties have resources to count over night and other counties don't.

Florida as an example, was counting mail-in/absentee ballots early. Pennsylvania conversely the law says mail-in/absentee ballots start being counted on election day. Both PA and GA report in-person votes first, which skewed heavily Republican, hence the early reports had Trump ahead. And only later did the mail-in ballot counts get reported, again typically in larger tranches than from sparsely populated rural areas.

The detail Mishkin, et al, supplied was the most interesting and useful part of any election coverage I've ever seen. The 99% of remaining coverage is next to useless. These folks went into quite a lot more detail on Arizona, than any other state, because they called it days before any other decision desk and were getting a lot of push back. The tranches of ballots are not heterogeneous, they're not random. They get bundled by type (mail-in, in-person, provisional) by precinct by county. And the statistics folks have a range for how many they expect for each, and which way and how much they will skew. And behind the scene they are getting even more detailed information on where these tranches came from, what's outstanding, and they can estimate how the result will track. At no time did the stats wonks seem at all surprised by what was going on, even though the talking heads were sometimes surprised.