A few more words, and a second thought, on Mark Kleiman's express-voting idea. Mark wonders at the possibility that it could be a denial of equal protection to speed up everybody's voting (thereby allowing more even of the non-express-voters to vote) instead of preventing many thousands of people from voting at all because there are too few fancy new voting machines.
"Note that the proposal would advantage, rather than disadvantaging, voters taking the slow option. Their wait in line becomes shorter as large numbers of other voters were diverted to the fast line. Clearly, the people choosing the fast line who were slow voters would be disadvantaged compared to the people choosing the fast line who were fast voters, but each of them (though not all of them) would have the option of avoiding that disadvantage by waiting in the slow line.... But if it's a denial of equal protection to allow some voters to vote in return for their promise to vote quickly, what is it to deny hundreds of thousands of people the opportunity to vote at all? I think the technical term is 'Straining at gnats and swallowing camels.'"
Under the circumstances I have a lot of sympathy for Mark's proposal. But I still think there's a problem with this particular gnat. Denying hundreds of thousands of more or less randomly selected persons the opportunity to vote at all is going to raise fewer equal protection/ VRA questions than will a procedure that makes it more likely that any given person will be able to vote-- but allows a bigger increase in that chance for literate English speakers and straight-ticket voters than for others. In a democracy with a different history I think literacy tests for voting would be a fine idea. In our democracy, they're considered particularly constitutionally odious, and for good reason. (Someday, perhaps, Jim Crow will be far enough in the past that literacy could be reintroduced as a relevant consideration; but we're not there yet.) And equal protection/ VRA questions demand that one look at relative changes, not only absolute ones.
So if this were a Broward-only-election, I think that would be the end of the story; express-voting would be impermissible. The catch is that it's not. If we were to resurrect the avowedly one-time-only equal protection claims made in Bush v Gore, then the variation across counties might outweigh the variation within Broward. That is, the unfairness to Browardites relative to the rest of the state could count for more than the unfairness to less-literate Browardites relative to more-literate ones. If Broward is more black or Hispanic than the rest of the state (which I assume but don't know) then the VRA might push the same way, meaning that we wouldn't need the Bush v Gore argument at all.
So, on reconsideration, I think it's at least legally plausible that express-voting might be permissible. Administrative personnel and government officials have an obligation not to knowingly act in violation of the constitution, even if what they want to do is reasonable and democratic, and even if a court wouldn't be able to stop them in time. (That, I take it, is InstaPundit's point.) But if there's room for legitimate legal and constitutional doubt, the officials' own best understanding of the constitution can allow them to try things, even if a court might later disagree with them. And, having had an extra day to think about it, I now think there would be such doubt. That means that I think I side with Mark; this would be a reasonable thing for Broward officials to attempt, under the circumstances. The best interpretation of equal protection and the VRA might rule the experiment out; but this is not so certain as to require constitutional bad faith on the part of those who might attempt it. [Final note: Mark implies that I'm a law professor, which I am only in the minor way that I've taught some courses that were cross-listed with law-- constitutional theory, philosophy of law. I don't have a law degree, and I don't have an appointment in the law school, only in political science. I'm a serious student of constitutional law, but not really a scholar of it.]