Thoughts
This is an extremely cynical take, but
for Kling to drop the Serenity target is like, you've already failed. You've already had to cut back scope. You can't brag about how you're tackling this project with a huge scope and then cut the scope.
Maybe it makes sense. I don't know.
LadyBird is certainly unique, and it remains unique, and Kling remains a talented developer.
One of the things that used to make it unique was the build-everything-yourself approach. If that's gone because you need to pull in third-party libraries, that makes it less unique. (One of the stated reasons for not targeting Serenity is that you can't pull in third party code with Serenity.) I think it could work (well I think it could fail in the way that it is currently failing), if you pulled in stuff like glibc or standard compression libraries. That's stuff you can't do in Serenity.
I've talked about this before, but Kling saying 'we support svg' when they support a very small subset of SVG leaves a bad taste in my mouth. That 'good enough' attitude works when you're doing your own thing. I am a fan of Kling in general and of SerenityOS. But I think Kling's attitude works for SerenityOS because Serenity isn't compatible with existing OSes. It's something different. Gemini proved to me that "something different" could work for the web too, to some degree. (Obviously Gemini doesn't work in some ways.)
In my mind, compatibility is always the number one criteria for a web engine. Which is why I've said building a new one is an impossible task.
Like my last LadyBird post, this is super rambly.