There are a number of studies that deal with that in a couple of different areas, and you're right, there are a lot of ways that people can vote. The 2003 study of the 2000 election by Larry LeDuc and Jon Pammett threw a lot of these things into a category they call “personal administrative”. One of the challenges with many of the additional opportunities you talk about—to vote by mail, to go to the returning office, that kind of stuff—is that there's a very high personal cost to voting in that fashion. So somebody has to be quite interested in doing that. Increasing numbers, they are. But overall, the turnout continues to decline.
So when you look at the two significant changes here, perhaps the most significant is the creation of voting opportunities on a Sunday, and the second is that with the Sunday immediately before election day, it's an area that's in very close proximity to where you live—essentially the same location as where you would vote on the regular election day—and that really reduces those personal transaction costs that make it difficult for people to vote, which become a burden.
About the Sunday voting, which is absent right now, I talked about studies where they think, based on evidence elsewhere, that would increase voter participation by 10%. It's interesting to note that just in general, in terms of Sunday voting, we're actually the exception in not having it, and adding it for advance polls won't change the fact that we're still the exception. But of the OECD countries, I think there are maybe half a dozen that do their voting on days other than Sunday, and we fall into that category. There is a consistent trend through a lot of the research and a lot of the studies that says Sunday voting would make a big difference and would have a positive result.