The NBC analysts I watched last night believed that, in some states and very approximately, Joe Biden got about the same percentage of votes as Hillary Clinton did, and on top of that some of the 3% of voters that switched from voting independent. 2016, the "also ran" candidates commanded approximately 5% of the vote, and this year it looks like under 2%, with a generally higher turnout obviously.
So the question here is, why would anyone doubt that the turnout was higher this year?
The amount of money spent on US elections eclipses the annual total economic output of some small countries. The total spending by candidates, political parties and independent campaign groups in the 2016 race was US$6.5 billion – comparable to the GDP that year of Monaco, Kosovo or Liechtenstein, and more than double that of Liberia.
The 2020 election cycle is forecast to smash previous spending records, with the Center for Responsive Politics estimating it will cost US$11 billion. That would be comparable to the 2019 GDP of Equatorial Guinea or Chad.
Campaign spending was up, so it shouldn't surprise anyone that all that money managed to mobilize more voters.