More people would probably use mail-in ballots through USPS if they didn't have to pay for postage.

Two obvious, unaddressed problems in how the county Registrar of Voters approaches this whole topic:

1. "Registrar's office was prepared for Election Day," yet "The issue was the 150,000 vote by mail that all came in" that day [said registrar Shannon Bushey]. Ever since election day, the Registrar's office has made similar statements, citing dropped-off "mail" ballots as if they were a surprise, or quirk. That's wrongheaded. They were 37% of ballots cast, a trend evident for YEARS. They decided some races. The Registrar was NOT "prepared for Election Day" if its rituals treated 37% of ballots as an afterthought. Since they take up most of the ballot-counting time, why does the Registrar not prioritize them more -- collect them from precincts as they arrive, start counting them immediately, rather than only after polls close?

2. The Registrar stubbornly clings to absurd Orwellian pretenses: "she had alerted the board and the public that the county's vote total likely would be ready by about 4 a.m. [November 5] - the final results were posted online at 4:36 a.m. -- consistent with previous elections." WRONG. Nothing whatever was "final" November 5. The Registrar's office had only begun counting the 37% of dropped-off ballots, which took nearly a week, followed by 14,000 or so "provisional" ballots needing further checking. Yet the website used language like "semi-final" and "precincts completely reported," and the Registrar speaks with a straight face about "the county's final vote" being posted Nov. 5 -- when over a third of the ballots were known to be uncounted. How about we drop that nonsense?

Why is everyone freaking out that it took several hours to get ballot counts? Because the Mercury News press deadline wasn't met? Because people didn't know when they went to sleep and instead knew first thing when they woke up?

Quick, inexpensive, accurate - Pick two. THey choose inexpensive by doing vote by mail which this article says is beyond what any other urban county has done. The remaining choice is that you can either have the answer fast or you can have an accurate answer.

If we went 100% vote by mail then it would take even longer to count votes because the signature verification would take more longer with more signatures to check.

"The Registrar's office was prepared for Election Day and had by Nov. 3, counted all of the posted mail-in ballots it had received by the Oct. 31 deadline, Bushey said."

How many votes was that? [See calculations below: 293,000-82,000-150,000=about 61,000 early votes, including ballots received between Oct. 31 and Nov. 4.]
How many ballots were received after Oct. 31, and before Nov. 4? When did they get counted?
How many votes were cast at precincts? [about 82.000] And these are paper ballots like the mail in ballots? Were they counted before or after the 150,000?
"150,000 vote by mail" were left at precincts and transported to the office after the polls closed.

"Supervisor Mike Wasserman noted that about 51 percent of county's approximately 575,000 voters cast ballots [about 293,000 actual votes] in the Nov. 4 election and that about 72 percent of county voters now request mail-in ballots [about 211,000 {72% of about 293.000}] leaving only 28 percent [about 82,000 {28% of about 293,000}] casting votes at polling booths."

Soubds to me like the greatest amount of time was spent manually checking mail-in ballots against signatures on file.