I think those numbers Stephen posted are dramatically off, for a couple of reasons.
Reason #1: Like James pointed out, ranking adjusts by sales velocity, not just raw sales. So if your book has been hovering at say a 10k ranking for the last two weeks, you can *roughly* compare it to another book that's been hovering at the same ranking for an extended period of time. But most books don't sit still like that, which makes it very hard to judge.
Reason #2: Those numbers change, a TON, based on current velocity of sales. For instance, in late Feb or early March Locke reported that the #1 seat resulted in over 5000 sales a day, every day, for the period he held it. Does #1 rank still do that much, or was that just his book? Does it do MORE now, because more people are reading ebooks than were in March? One book at the top seat might be selling 10k a day; another might be selling only 3k a day, and bump down fast. Really hard to say.
Reason #3: The numbers all seem very low to me. In fact, across the board they are lower than every reported ebook sale I've seen, including my own. A 10k rank seems to be something more like 10 books a day, right now, and rank 1000-ish has been holding at about 50-60 books a day for a few weeks now (gradual upward trend). That's based on real observations of books actually for sale, by myself or other writers I correspond with on the subject. But again - that's not perfect - because it's about velocity, and rank decays over time, so a big one day boost of 500 sales might kick you up very high, and even at one sale a day afterward, you will still only slide down in the rankings over a period of weeks.