That's not really a solution, is it? The issue is not about coding for IE6, it is about understanding how the various vendors' products act.
As for not supporting IE6, shall we drop IE7 also? The same root cause of most IE6 woes is extant in IE7. If we fix something in one, we often automagically fix it in the other.
As for numbers, don't confuse market share with visitors. Just 1% of the US users is still 22.7M people. (227,719,000 total users as of 8/09 per Nielsen Online) What's your marketing demographic? Businesses that still run IE6 because of an inventory of internal web-apps that depend on it. What about Joe Sixpack and Helen Housewife who are running XP, and haven't upgraded because of whatever. They're happy with their browsing, email, and word processor. Seniors are the fastest growing segment of the web market. How many of them are using computers handed down from their children, and still running the original software?
The IE6 bugs and their fixes or work-arounds are well known. There is no good reason to deliver a broken page to IE6. It need not have all the bells and whistles available to more modern browsers, but a professional knowledge of its limitations will allow you to code for a graceful degradation without extraordinary effort.