The short response would be the "web browser" started to fill the role of "cross platform on demand app sandbox" instead of "plain hypertext document viewer" and that will forever be evolving to match the capabilities of devices themselves. More than any other contributing reason to the churn (including the highly charged/passionate reasons), if you defined a "web browser" to be a limited subset of the functionality the user desire for a constantly evolving then an "on demand app sandbox" app with a different name would still exist and it'd just be everyone using that application instead of a web browser.

Because the scope keeps growing as well. People have been pushing desktop apps to the browser for years.

I wish the browser was limited to showing websites, and that technology for desktop apps evolved. But that's not how it goes for some reason. Some kind of imperialism in platforms? Systemd wants to take it all, browsers want to take it all, ...

Web standards are moving fairly fast, and modern browsers are getting more and more complex in the process. It would likely be extremely hard, if not outright impossible, to build a browser completely from scratch today because of the sheer amount of moving parts.

Browsers sort of went from a network document viewer with forms, to a whole application platform. It's sort of like grafting on changes to Acrobat Reader so it becomes a PDF viewer with WinForms + DirectX.

The fact that we expect basic text-and-image websites to display along side full online 3D games using GPU acceleration and might even interface with a few web assembly modules... is sort of insane. It's all just an application platform now.

I think that just means that browser complexity grows with the complexity of software globally.

Definitely not a dumb question. It's a really good question that I rhetorically ask myself every time I start one. And I'm glad there have been some good answers to your question here. Though for me, it will never be answered , because I helplessly detest the subject.

It's simple - interfere with the competition, useless engineer job security, performative "working" , and bad incentives all around. They're on a gravy train and don't want the ad revenue to ever stop. A truly independent browser threatens that entire ecosystem, and that in turn results in collusion and patterns of behavior intended to make things nearly impossible to keep up with, even though the actually functional bits that consumers give a shit about haven't really changed all that much in the last decade. What Microsoft did with IE, all the major browser developers are doing now to reinforce their anti-competitive collective behavior and each can seek as much rent as possible from their respective walled gardens.

Not a (consumate, formal trained) developer so please don't take my word on this, but for what I've read the big issue development wise about forking Firefox is that its codebase is kind of a mess, in like dissociating the Gecko engine and the UI would imply a tedious, herculean effort, and this is the main reason there are few Firefox "forks" (librewolf, etc.) with little modifications respect to Firefox, whereas there are tons of forks of Chromium.

Now, I understand making a Firefox fork that doesn't include this new changes in upstream doesn't imply major changes in its code, but still it seems like if you want people to use your fork requires more than that, or maybe you'll end up with a scenario like Debian vs Biebian and Hannah Montana Linux.