Doesn't that seem like a scheme to allow for fraud? If Precint_A has 1000 registered voters, and 700 of them mail in their ballots, then turnout in _A cannot be greater than 30%. If among the remaining 300 half voted in person, then turnout would be 15%. Ergo, we would expect to see very low turnout in most precincts under the theory that most people mailed in their ballots. Meanwhile, in AVCB_1, ballots are mixed together from many precincts (see in the document you linked: "A jurisdiction that uses AV counting boards must establish an AV counting board for each precinct. (Jurisdictions of 250 or more precincts are exempt.)", so there are no numbers of registered voters available to constrain the maximum possible number of registered voters, so any number of votes can be counted in AVCB_1, whether real or fraudulent, and if there is no way to trace each ballot back to a legitimate registered voter, then there is no legitimacy to the vote counts at all.
Doesn't that seem like a scheme to allow for fraud? If Precint_A has 1000 registered voters, and 700 of them mail in their ballots, then turnout in _A cannot be greater than 30%. If among the remaining 300 half voted in person, then turnout would be 15%. Ergo, we would expect to see very low turnout in most precincts under the theory that most people mailed in their ballots. Meanwhile, in AVCB_1, ballots are mixed together from many precincts (see in the document you linked: "A jurisdiction that uses AV counting boards must establish an AV counting board for each precinct. (Jurisdictions of 250 or more precincts are exempt.)", so there are no numbers of registered voters available to constrain the maximum possible number of registered voters, so any number of votes can be counted in AVCB_1, whether real or fraudulent, and if there is no way to trace each ballot back to a legitimate registered voter, then there is no legitimacy to the vote counts at all.
Does that make sense to you?