Wouldn't it be better to offer a legacy addon API shim - old addons would still work even on new Firefox.
Look, I get it. You have your extensions and you don't want to see them go away. But you're a tiny niche and even other power users don't use these extensions that you find essential. Certainly the vast majority of users rarely install extensions at all. Mozilla is doing the right thing and an entire thread praising the performance of the latest version vindicates their approach.
Where is the next Servo based browser? Instead of refactoring the 20year old mess (with COM and XML), the should focus 99% on Rust and Servo, and finally release a stable build with a lightweight HTML5 UI (Vivaldi).
Well, you answered your own question. Clearly they have a userbase that will stay with them out of inertia or loyalty. These users can now benefit from the improvements in the latest release. If they released a new browser, few people if any would try it out because of inertia. The lack of users would kill the product pretty quickly.
Of course, some users who rely on old addons will be left behind, but they have the option to use Firefox Extended Support Release, which will get security updates. Eventually the extensions situation will sort itself out and these few users can move to the main release.
The "legacy" Firefox extension API is powerful because it reaches its tentacles into every corner of the internals of the browser. Keeping that API stable involves freezing a large amount of internal architecture and behavior.
Firefox needs to evolve if it's going to live, and it can't evolve in the ways that it needs to if so much of its internals are effectively frozen. It's a painful drag on development, slowing things down and preventing necessary changes.
I'm not even going to get into the problems with the API itself except to say that it had many, including widespread dependencies on synchronous behavior.
Mozilla's sacred-cow-level commitment to preserving legacy extension support is a big reason why Chrome beat Firefox on so many technical levels for so long. It ate up a ton of developer time, held off or killed critical architectural changes, was the cause of a lot of crashes, hangs, memory leaks, security problems... Thank goodness that's about to be over.
I'm sorry that you'll miss it but this is a really important step. Mozilla is doing a great job with WebExtensions, I suspect you'll be happy with how things shake out in the end.