That's not how it works.
You're making an assumption that a developer will devote the same amount of time and enthusiasm to any project, and therefore any "Me-too" development is a loss to an established project.
That's more or less how it works in commercial development - you have a developer working 8 hours a day on whatever project you give him. It's not how it works in FOSS, which is actually a selfish development model: you work on whatever you want, as much as you want.
It's better to have someone sit down & bang out a yet another variant of X than to have them not turn out anything at all. The new version of X might be enough of an improvement to attract other developers and get you a new killer product. Not doing anything will result in nothing.
He wants to build a lightweight KDE, so he goes off and writes some code and with any luck creates a working KLyDE WM. If he tried to do this within KDE itself, he'd hit a load of politics and entrenched views and spend all his time arguing. He'd get nothing done. Nobody wins.
What you see as duplication and waste is a vital and useful part of FOSS development. Your recommended alternative would result in stagnation as people stop working on what interests them to fight against each other to drive existing projects towards their vision.
That's what the harm would be.
Linux is also massively more widely-used on servers than Windows, but remains far more secure.
Life imitates art? http://dilbert.com/strips/comic/1997-03-29/
The most important one is that it can gaurantee that the software you're running is the software you THINK you're running.
Simple example: Someone nasty gets access to your Linux box and installs a rootkit. This includes a modified version of "ps" that won't show the rootkit process(es), making it harder for you to notice it's there.
If you use a Linux machine that's set up to take advantage of the hardware, you could have it set to, say, only allow software that was signed by Canonical to run on it. This would mean that all your Ubuntu software would work fine, but the new version of 'ps' our malware installed, that wouldn't run. This would alert you instantly to the fact that someone has installed malicious software, and allow you to get rid of it.
Of course, we all use more than just Ubuntu's own packages, so you'd also want it to allow software that you personally signed with a secure key (i.e. one that ISN'T kept on the machine so the bad guy can snag it) - otherwise you wind up in an Ubuntu "walled garden". So it's important that you can say who you trust to provide software you can run on your machine.
It's actually a nice idea, (I'd certainly like my next PC to take advantage of it) but unfortunately one that gets constantly overrun with paranoid hype about it being designed to kill FOSS.
Really?
You might want to rethink that "I have freedom" opinion a bit..
Because the typical end user is stupid and forgets their password.
On a normal laptop, this means a bit of inconvenience.
On an encrypted laptop, this means a loss of all data.
You have to have solutions for this problem in place before you can roll it out.
The reason configurability matters so much isn't that we want to change a million pointless bits of eyecandy.
It's because there are certain features we want that not everyone else does.
Simple example: When I Alt-Tab to a different window, I *require* my mouse pointer to be moved to that window as well. This is a feature whose absence drives me *nuts* - It's literally a deal-breaker for me not to have this feature available.
Other people hate their mouse being moved by a keyboard shortcut. I can understand that, whilst not agreeing with it. So the only way a WM can keep us both happy is to make this a configurable option.
When I use a dual-screen setup, I *hate* the Alt-Tab list showing me the options for both screens - I only want to be able to switch between the windows in that one particular screen. Other people want to be able to switch to any window in either screen. Still others want to be able to switch to any window on any desktop.
That's quite a range of desired behaviour just for something as simple as the alt-tab function. Not having it set to the way they like is a big problem for people who spend eight hours a day trying to Get Stuff Done. Thus the only way to make a WM that everyone can use is to make it very configurable. Not so people can get endless special effects and fiddle with window decorations; but to get the behaviour you want and expect.
I use FVWM personally, and I once worked out that the functionality I had built in to my hotkeys and preferences was worth an hour of productivity a day, just in the time saved on mundane, repetitive tasks. Configurability matters. It matters a lot.
It's why the only books on my reader are from either Project Gutenburg or O'Reilly - all are DRM-free.
It'd be highly amusing to me if, in a few years time, Windows users are keeping a copy of Linux around because "I need it for the games"
As somebody who hasn't bought (or pirated) any games in about a decade (other than a few of the Wii Lego series) I have to say that the only downside of boycotting all modern games is that you have to find something to do with all the extra free time and money.
My performance is simply marked down as "godlike" for all categories and we're done.
1. KDE feels even more bloated than Gnome.
Note - I said "bloated" not "slow" - I don't care how fast either of them are, there's too much of them both. They get in the way.
2. KDE doesn't do what I want any better than Unity.
Example: Here's one feature I want that even Windows fricken VISTA gets right, but that Gnome/Unity/KDE et al make so difficult my choice of WMs is limited to FVWM2 or xmonad: Mouse follows focus. It's such a simple thing. When I move my mouse to another window, I want that window to gain focus. That's easy. But when I alt-tab to a different window, I want my mouse to go to that window too. I don't want to be in a situation where one window is accepting input from the mouse whilst another one takes input from the keybaord. It's moronic.
3. I can't be bothered
KDE has lots of impressive features. I don't need or want most of them; I don't have time to wade through all the tutorials and howtos to find out what the few useful things it has are. I'm not interested in investing the time to learn yet ANOTHER interface.
At home I use Gnome because it's the default and it's a PITA to set up lightweight WMs to work with my two different-sized screens (laptop + monitor). I typically do little other than browse the web from home, so it's not enough of an issue to do anything about. At work I use FVWM2 because I've set it up to work EXACTLY the way I want it to and it does everything I need. Though I am considering a switch to Xmonad for a variety of reasons.
The only thing I need a WM to do is give me a place to run command-line shells and the few GUIs I'm developing for. KDE and Gnome both suck abysmally at giving me this. That's fine, I'm not their target audience. Trouble is, they both suck at being as comprehensive and user-friendly as OS X or Windows. We'd be better off if they were both scrapped and a new DE devised, one that didn't have the "for hackers" legacy deep in its foundations.
The more unsuccessful an interest group becomes, the more strident and extreme its demands become.
This is a direct result of the failure of ACTA, SOPA et al - a desperation move, not the head-in-the-sand reaction implied by the summary.
So keep an older version of Firefox installed to run older versions of Java with. Why would you want a cutting-edge browser to replicate legacy problems?
According to the latest official figures, 43% of all statistics are totally worthless.