Since there have been a plethora of "we're doomed!!!" diaries of late...

My advice is:  relax.  

• Pay attention to the polling, in particular the polling aggregates.  I like 538, some others prefer RCP.  Don’t sweat individual polls that look bad, outliers happen and are a normal part of polling.  (If a polling firm DOESN’T occasionally produce outliers, that may be a sign they are cooking the books).  By the same token, don’t boast about outliers in the other direction.  USC/Dornsife released a poll today with Biden +15 over Trump.  Do I believe that?  Not at this time.  But into the hopper it goes.
• Pay attention to “generic ballot” polling, and this election cycle, polling on BLM.  Depending on who you ask, the generic ballot is anywhere from D+6  (RealClearPolitics) to D+10 (Civiqs).  Excluding a laughable push-poll from McLaughlin which is D+1 (the poll includes questions like “Should President Trump continue to fight to change Washingon?” and “All things being equal, do you prefer free-market capitalism or Big Government Socialism”—loaded questions no reputable pollster would ask), 538 seems to be around D+8.  
• Relevant to this electoral cycle, support for Black Lives Matter seems to be steady around 50% support, 38% opposed, with a tenth of the country undecided.  This is polled less frequently, but it’s a good barometer of how well the GOP’s gambit of “The Dems are going to let big-city rioters wreck the country” is going to play.
• And speaking of that gambit—the GOP and its media organs and sympathizers are doing a full-court press, trying to plant the seed that Democrats are unfit to govern because rioting in a few major cities wasn’t swiftly quelled.  And—this is important--they’re trying to make this a self-fulling prophecy.  Part of trying to convince people that X is true is convincing them that other people think X is true; many folks will follow the herd.  
• But remember.  While not all politics is local, much of it is.  The generic ballot is not a good predictor of any individual race, even if it is a good predictor of the aggregate results.  Riots in Wisconsin are more likely to affect that state than they will affect voters in Idaho, Hawaii, or Arizona.  
• Pay little attention to anecdotal evidence.  I’ve heard all the stories of “shy Trump voters”.  I’ve never met one, though I know some loud Trump voters who insist that people who voted for Hillary in 2016 have confided support for Trump, usually with the riots in PDX cited as a reason.  See the point above about self-fulfilling prophecies.  This evening, someone wrote a diary claiming that Wisconsin was turning against Biden hard as a result of the Kenosha riots—but that diary didn’t have any links to any actual evidence of such; it was something that the diarist had apparently heard.   Data is king.  Anecdotes and hearsay aren’t data.
• And finally:  GOTV.  And vote.  One third of politicking is convincing persuadable voters to switch.  Another third is motivating your own voters to the polls.  A final third is trying to demoralize the other side’s base, to get them to stay home.  A key figure to watch out for in Trump’s approval is about 42-43%.   That’s about the size of his base.  If he goes above that, he’s starting to flip independents.  If he drops below that, then his own base is starting to be demoralized.  Right now he’s right about 42% in the 538 average, and has been stuck there about a week.  He won’t win a fair election with those numbers.