Newer IE browser have 2 problems. First, they still have major problems. eg. I was writing a fairly low-level JS library (not building on any existing libraries) and had to put a work around in for IE9 because setting the "type" attribute (like the one in <link type="text/css" ... ..../>) is not standard -- apparently it's read only despite this being a writable element since DOM 1 (we're on DOM3 with DOM4 in the works IIRC).
The second problem is that MS doesn't provide enough incentive for users to keep their browsers up to date. There's no excuse for ANY system to be running IE6-7, yet developers must still write web-app code with them in mind.
The fact is that MS got scared by Netscape, so they poured development into IE. After they crushed Netscape and were finished with IE6 (an amazing browser at the time), they then dropped most of their IE staff because they believed .NET was going to replace the web (I guess Java's failure in this regard didn't teach them anything as they just copied java with c# and tried to do exactly what it did). After ajax happened and Firefox entered the picture, MS still failed to act. Safari and Chrome entered the scene and MS still failed to act (except for stalling the standards committees because MS was too lazy to change anything).
Now MS has realized that they don't have the market power to force their way (webkit is now the standard-breaking problem), they have decided that they must get with the program, but they are way behind the game. FF, Chrome, Safari, and Opera have a huge headstart and their teams are big enough that MS can't really out manpower them. Unlike Netscape, they also aren't going broke because they don't have a business model (the Netscape model was to give the browser away for a while and then start selling it after it caught on -- what little potential there was disappeared when IE was offered for free, so Netscape went under).
Very few people use IE10 and I suspect that IE11 adoption will be similarly slow. With the popularity of FF, Chrome, and Safari outweighing IE (especially when you consider mobile devices), few developers have time or inclination to deal with IE's oddities (which as I pointed out, still exist). Code for these non-IE browser and fallback for old IE versions. If newer IE versions happen to run the better code without breaking, great, but if not, the newer IE versions can just switch to quirks mode and give the users the old fallback version. This view may not be particularly good, but it's the one I see quite a lot. On the flip side, MS stated that they didn't intend to implement webGL (and similar), but perhaps being forced to the sidelines by developers was motivation enough to force them to stick to the standard (Thus we have webgl in IE11).