Obviously there are exceptions, and some tools are so advanced that it's necessary to be able to customize them. No argument from me there. But for most tools, and most users, there's just no hunger for customization. Almost nobody wants to have to manage an array of options to do a web search.
What is even more shitty is when soem feature is removed like the System Tray and first they pretend they do not understand what you mean when you say this feature is very important, then after someone wastes his time to explain a n=th time again how working people use the System Tray features at work or at home the GNOME dude finally admits the issue and offers some workarounds that are not equivalent but are "possible to do but with lot more work".
On short, some people do work on the computer, some people use search engines for work to find relevant stuff, this people do not ask features like "please use the exact same padding everywhere because I am OCD" or "please make those buttons/corners or edges smoother so I don't cut my tongue when I lick my screen", this people want the customization to do a task.
I am not sure why simple people have a SEGFAULT if they randomly end up in the Advanced section of a settings section, what do they expect when they open Advanced? Google main page has a small link for Advanced search see https://www.google.com/advanced_search?hl=ro&authuser=0 how many GNOME users got hurt by this link existing?
The Titanic didn't sink because of the arrangement of deck chairs, and Mozilla didn't sink because of any features they did or didn't provide (and GNOME didn't fail to achieve significant market share in the first place because of features either). The actual problem doesn't have anything to do with the stuff on the deck at all.
It's the one-two combination of vendor-lock in and bottomless marketing budgets. Since most of the value of the Windows platform and the Web is the immense amount of stuff that's built on top of it, there's a huge lock-in effect that prevents you from even reaching parity, much less exceeding it. And in order to overcome the marketing budget and pure inertia, you need to be ten times better, not just on par.
If GNOME becomes as usable as Windows, it won't have anything to do with what they actually do in the desktop environment itself one way or the other, whether it's continuing on the road they take now, or reverting everything back to the way GNOME 2 was, it's totally irrelevant. GNOME becoming usable will be entirely because of Valve investing in Wine, combined with a whole bunch of other apps moving to Electron and shipping Linux versions because heck why not?
Unfortunately, while they are probably already on par with Windows, they aren't ten times better than the Mac:
* The Mac has a bottomless marketing budget. Good luck competing with that, GNOME.
* They've shown a lot more restraint than Microsoft has, probably because macOS is considered a niche product to round out their catalog rather than being their one and only operating system like Windows is for Microsoft. They have even reversed course on a few anti-features, like adding back USB-A ports to the Macbook Pro even though it made the laptop slightly thicker. And unlike Windows RT, they didn't lock down the ARM Macs.
* Those tectonic shifts I mentioned that made Linux usable? They also make the Mac usable, because Wine is open source and Electron is basically its own operating system. Anything truly good that GNOME does, Apple can copy it just like Chrome copied all the really good stuff from Firefox.
What would work IMO is someone with money paying the developers and the designer but force them do do customer support, you don't play with the new shit until most tickets are resolved and customers waiting for response are satisfied. maybe a paid support would help too.
Market success has almost nothing to do with product quality. Well, okay, it does, but only in the sense that you need to not actually be a total fraud. You can get away with dismal quality as long as your marketing is good [1]. In formal terms, software development is a loser's game [2].
[1]: https://danluu.com/nothing-works/
[2]: https://news.ycombinator.com/item?id=26660679
This implies that the churn you're complaining about has nothing to do with market success. You might not like it, but that doesn't mean the failure of Linux on the desktop is actually caused by it.
"Hold on", you say, "professionals want options, like user-expandable RAM etc". No, that's the misconception about the concept of a "professional". Unless you are a hardware engineer, tinkering with your notebook's internals is the absolute opposite of professionalism, its either a completely misguided waste of time and money, or a perfectly fine hobby.
Real professionals get work done. Customizing their workspace is something they feel ashamed to do, because it's procrastination at best.
Just that exactly is the workspace changes a lot, the hardware internals are hardly ever a consideration for any professional, it's easier to just buy something that works well from the start (and hardware costs peanuts).
People want to accomplish stuff, this means they need tools to do stuff and if a tool can be customized to do the stuff faster or better people want the customization. At my job we have paying users that requests features that indeed are work related(not moving shit around). I know GNOME-minded people will disagree and they prefer to bend their work to fit a guru-s vision.,
Now my turn to ask, why do people like you think there is a generic and basic solution that works at the same time for the casual user and for the user that has a lot of tasks to accomplish? Is there some theorem that shows this, like "The GNOME theoreme of product design, keep removing features until the shit convergence to the local minimum where you find the minimum product and the minimum set of users possible.