Yeah, I thought you might have meant the 168.3 was total votes cast in that election, not total eligible voters. Taking 66.6% participation to be ~2/3, then total eligible voters should be 3 x 168.3/2 = 252.5 million or a bit more than 3/4 of the total population. That in itself gets my attention as it would mean only 1/4 of the US population is under 18, which seems a bit on the low side to be healthy
Percentage of those registered who actually voted went up 3+ points, too - 89.6% for 2008 and 91.8% for 2020. I might try to get the numbers of total voters in senate races to see how well those totals match the presidential vote totals. One indicator of suspicious voting patterns would be a high number of ballots with only the presidential vote marked
As far as the time limit on editing, I think it has gotten shorter. It used to be a couple of days and now it is definitely a lot less than 24 hours
