Mozillas browser.html project may be helpful too, but it seems to be a bit too experimental for now.
For an extension to get support, we need to rank it as in-demand, and then try adding it to Brave to see what missing chrome.* APIs it might need. https://twitter.com/bravesampson is leading this charge, using https://github.com/brave/browser-laptop/projects/1 to keep track and engage with developers.
For future extension APIs that go beyond the chromium ones, we won't support XUL, but I'd like to take advantage of our Muon (https://github.com/brave/muon - we forked Electron, because security) React.js-based front end.
In 2Q2017 I'll hope to have more to say about extensions and where we'll go. Interested extension developers are welcome now to join our https://community.brave.com/ discourse. Thanks.
I was not talking about the type of extension that allow to do some common things based on small api created specifically for that.
I am more interested in extensions that get full access to existing react components, and can modify the browser ui in arbitrary ways, e.g. by adding splitview similar to what cloudy has, or adding proper tabs on mobile instead of carousel.
Brave UX will evolve but slowly and without big changes that break too many users (Australis). Still, we need to be able to change our UX. So if we have powerful extension APIs into that UX, we can't promise never to break extensions that use those APIs in ways that become unsupportable.
All we can do is try to co-evolve nicely and cooperate better. I don't see a silver bullet here, but I do see how Mozilla has burned a lot of add-on developer good will -- and we won't make that mistake at Brave.