It'll go well for a while (and users even might love you for it for a year or two, since you can focus your effort on fixing bugs and not introduce new ones with new features), and then you'll slowly but steadily lose users to other browser that do adapt.
Losing users for a web browser, means losing search referral revenue in the short term (literally Mozilla's lifeblood), and losing web developers in the long term, which will break the experience even further.
Just one example that almost made me switch browsers: Web site translation. I was regularly using Chrome in parallel for that, but now Firefox fortunately supports it too (and in a privacy-preserving local way at that – a true innovation), so they keep me as a user.
- End-to-end encrypted tab, history, and bookmark sync across devices
- Content translation, as mentioned above ideally in a privacy-preserving way
- Plugin support
- Cookie-jar-per-tab and proxy-per-tab support (Firefox allows doing both through Multi-Account Containers)
I'm sure some other users also feel the same.
I'm impressed you don't even need a HTML renderer component and all that comes with it.
Also, it is possible to continue iteratively improving a complex piece of software like a browser, beyond just the web standards race. Security, privacy, performance, and reliability/bugs/code correctness are areas where new computer science is constantly coming down the pipe and worth integrating. And maybe other things like AI or features for AR/VR systems, though it's more debatable whether those belong in general purpose browsers.
Browsers aren't Unix utilities that should do one thing and just one thing and thus can theoretically reach a state of "done". But even there it's not always a certainty. For example, "sudo" being superseded by a simpler more secure "doas", and then more recently by SystemD "run0". Even simple utilities continue to evolve.
