"make it sure that it works on ie6" is the worst line you can hear at work. especially if you're developing a web app. even global websites like fb don't support ie6 anymore.
As with anything, it's a cost/benefit analysis. If you ran a company that sells a billon dollars of products online per year, would you ask your web developer to spend hours ensuring the site looks and functions decently in IE6 if that means an extra $26 million? I sure as heck would.
On the flip side, if you're short on resources and that time would be better spent adding functionality which would result in > 2.6% return, then by all means ignore IE6.
The problem is that their users are not a representative sampling of the overall browser market since they heavily skew towards web developers. If you are also making a site for web developers, then the w3schools numbers are valid. Otherwise, no.
http://www.w3schools.com/browsers/browsers_explorer.asp - IE6 has 2.6%, I don't really see a point in spending hours optimizing stuff for it.