Regarding the lead time and its effect on Worldcons and Worldcon site selection, Deb Geisler says it better than I can. So does Cheryl Morgan.
To my knowledge, a three-year lead time would not necessarily have improved Seattle's ability to commit to facilities. Their committee has not said specifically that a lack of sufficient lead time was the issue, and I am unsure that it was. It may have been, but there may have been other things involved as well.
In 2002, we still selected sites three years in advance. (The election was at Aussiecon Three in Melbourne, 1999.) Both of the original 2002 Worldcon bids -- Seattle and San Francisco -- were obliged to fold for reasons that actually didn't have anything to do with other groups dropping money on their prospective sites. In the former, the person controlling the key hotels decided he didn't want the business, and in the latter, the headquarters hotel priced the bid away, chasing more-expensive business (that they didn't get -- I was happy to hear that the "Jukebox" was empty on Labor Day weekend 2002). The Bay Area was lucky -- we had another Worldcon-capable site fifty miles to the south that was eager for our business. Seattle wasn't so lucky and had to fold.