They both have advatanges, and they both (yes both) could learn from each other. If I was to create an OS, it would have nothing to do with either of them. I say that Windows, and Linux's designs (functionality, GUI design etc) are the reasons they are so popular today, it's also the reason why they're not moving and people are still trying to make the "perfect" OS. Sure it's one heck of a project to take on, not given to everybody to make a pot of coffee and say to himself/herself hey, let's make an OS today. And no I'm not saying I'm one of those that could, I am saying I'm one of those that could either. I don't know, so I won't answer that question. Maybe I can, maybe I can't
.I know what you are trying to do, make the guy stick to his convictions, and test his belief in his own project, that's good in a way. but while you're there testing his faith in his own project, you're also stopping him from moving it forward at the rate it could move forward. To me, any effort to to make an OS is good, but any effort to make an OS totally different than what we all know and use today is even better because as much as some might hate to say it, windows (any version), Linux/Unix (any version and any GUI) and OS/2 aren't perfect. There better than what was available before, but there's much much room for some serious improvements. SOme that can't be implanted simply because of the way the code to these OS's aren't designed to be changed if you get my drift. Linux included.
So that's my opinion, make a 95 clone or an NT clone that works in Linux to, excellent, I'm sure people will want to use it once it's done....but to me, it's high time for something totally different, new, never before seen to start happening.
....but indeed if everyone spent their time answering questions no project would ever move forward. To me, energy is better spent at development rather than justifying 
