New VCU Poll Has McAuliffe, Ayala, Herring Ahead, But Take This One With a HUGE Grain/Pillar of Salt

We seem to be getting more polling these days on the Virginia 2021 elections. Just out is a new one from the L. Douglas Wilder School of Government and Public Affairs at Virginia Commonwealth University. See below for the key findings:

• “43% of likely voters would vote for Terry McAuliffe while 34% would vote for Glenn Youngkin…This widens McAuliffe’s lead over Youngkin from August when he led by a much slimmer margin (40% to 37%).” Why would McAuliffe’s lead have increased since August? Who knows.
• “In the governor’s race, 23% of voters remain undecided or unwilling to vote for either candidate – the same percentage as August.” This is strange…a much larger percentage of undecided voters than we’re seeing in other polls, and also unchanged since August? Hard to believe.
• This is also odd (33%-30%, with another third undecided or saying they wouldn’t vote for either candidate, at this point in the election is VERY heard to believe): “Ayala lost some ground to Sears in the race for lieutenant governor in the past month, and her lead is now within the margin of error (33% to 30%). However, 20% of voters said that they would not vote for either candidate if the election were held today while 16% remain undecided.” These numbers make no sense, basically.
• As is this: “Likewise, Herring lost ground to Miyares since the August poll. Herring lead by 11 points in August, while the current poll shows that his lead has dropped to 6 points (39% to 33%).”  These numbers also make little if any sense.
• “Poll respondents were also asked which party they would rather see in control of the Virginia General Assembly. All 100 House of Delegates seats are up for election in
  November, with Democrats currently holding a 55-45 advantage. Voters remained split
  on which party they preferred to control the chamber. Democrats had a slight edge over Republicans (43% vs. 39%, respectively), with 5% of voters undecided.” 

Also worth noting is that the writeup on this poll is very odd. For instance, both Hala Ayala and Winsome Sears are both referred to incorrectly as “his” instead of “hers.” Also, what pollster writes that their poll has a “margin of error of 5.35%” or “6.93%” (to two decimal places? LOL). And again, the undecided numbers in this poll are waaaaayyyy too high for this point in the race. Also, there’s no way that McAuliffe gained 6 points since the last poll, while both Ayala and Herring fell. None of that makes any sense, frankly. So basically, take this poll for what it’s worth, namely with a huge grain – pillar? – of salt.