Is it just me or did not a single one of those "l33t haxker shells" ever produceable a single ui innovation that lasted?
I mean, I remember there being a whole ton of wildly customized windows shells with menus and floating terminals and so on, but not a single thing stuck around?
I'm not sure about the Windows scene, since I only toyed around with one or two shells nearly twenty years ago, but the motivations for creating windows managers in Unix varied. Quite often they were about the appearance, customization, ease of customization, lack of customization, or low resource usage. I suspect that most of them were made for the learning experience or simply as a form of self-expression. They were never really about innovation. When there was innovation, it was usually in the form of small things like how we size and position windows (e.g. think about how it is possible to tile windows in Windows these days).
Besides, the term innovation is used far too much with respect to software, in the sense that a lot of stuff can be traced back much further than the so-called innovators will suggest. Many ideas have deep roots, but it took several (often independent) attempts until the technology or its users were ready for it.
Well, I installed Litestep as default shell on a 80486 laptop with Windows (I think it was Windows 98 SE, debloated). It replaced explorer.exe and performed better, IIRC I had more RAM free. I gave the machine away eventually, and they probably ran Linux on it. But because the machine was for my mother, it had to be able to run MSIE.