I guess main points is:
- to get windows first class support for ember apps (to have weight in enterprise)
- different builds out of the box (for IE11 and modern browsers) - same for enterprise
- more work about core composablity like (optional router, rsvp, edata, controllers) - will help newcomers easyly adopt ember or start integration into existing apps.
- More "glue" guides about how to deal ember app builds with webpack (in case of migration to ember from existing react/vue/angular stack).
Also, it will be great to have A11y, Ember-inspector, Glimmer-VM roadmaps with bullets or kinda, it may be an point for desigion makers.
I see alot of work aroud, but, it's not written in roadmap anyware.
From my experience if it not written/documented anywhere, it does not exists. Team can easyly switch focus without strict roadmap.
It's not about "do things only pointed in roadmap", just to have writted all planned things, documents are editable, and if team has new plan or inputs, it's not big deal to move feature back to "backglog" or "need discussion" lists.
Also, it will be greate to have kinda "likes" for non important features, to prioritize them (comminity members can write posts why whey need this feature, and get more "likes" to this feture from readers)
2018 year most feature - focused, and it will be great to have same or higher results in 2019.
Sometimes feedback on pull requests from core team members too slow, it can take up to 3-4+ months. It's discouraging. After 3-4 months you may have no time to work on requested changes or forgot "context". I think we need kinda bot for pinging and working on pull-requests. (example - DefinitelyTyped repo)