I wouldn't call it comparing apples to oranges. I'd call it comparing red apples to green apples. Not identical, but very similar. And even if you use the argument of people using mail-in this year cause of covid, it's basically a statistical impossibility for 100% of mail-in votes that rock up miraculously at 3am to go to one candidate. If it's apples to oranges, then it's because they ramped up election fraud this year like never before, and many players were willing participants just because they're never Trumpers.
You also look at Trump vs Biden rallies, 50,000 vs 12 people rocking up and the fact that, just in the last week, about a million turned up to Washington DC to support Trump (and no one asked them to either, they did it willingly), and there is no way Biden could have won. You can argue that a bigger turnout in a rally doesn't necessarily equate to an election win, especially since people this year prefer to use self-isolation as an excuse to not go to a rally. Technically it isn't an indicator. After all, you don't vote until you actually vote, but barometer-wise it's highly telling, especially with such a big discrepancy in turnout. Sure, not every single person, dead or alive, who voted for Biden turned up to his rallies, but on the flip side not every single person who voted for Trump attended any of his rallies either. Believe it or not, there are pro-Trumpers who also adhere to social distancing. It isn't 100% pro-Biden are social distancers and 0% pro-Trump are social distancers, just like 100% mail-in ballots for Biden also isn't realistic.