"Recruitment of gays? How many of them are there, and how many bother to vote?"
It's right there in the article.
"In fact, exit polls showed that self-identified gay people made up 4% of the total electorate in the incomparably close election of 2000, and nearly one third of those homosexual voters cast their ballots for George W. Bush."
If this is actually right, that would be more than 4 million homosexuals voting.
And how well did Exit Polls do in predicting the winner of the election?
There aren't four million. And pollsters really shouldn't limit their sample to residents of San Francisco and then claim their results are indicative of the hole country.