Starling for haxe/openfl is a unofficial port, nor Daniel, Josh or Adobe are related to it (yet).
OpenFL relies, I think, mostly on lime for the rendering, which means that Starling support is constrained to the lime implementation.
If AIR includes new features that are not yet (or can't be supported) by a target platform in lime, the Starling port will need several fallback implementations, which means a ton of work, incompatibilities, maybe different features exposed by lime/target platforms, and it will become eventually a totally different engine compared to Starling.
Idk what's the current roadmap, but Starling 1.8 was released a while ago for OpenFl.
Daniel changed A LOT in v2.0, he refactored Starling's core (Mesh, Styles, etc).
Maybe eventually, OpenFL (or someone else) will port it. But take into account future inconsistencies.
I doubt that Adobe give support to Haxe in the future, as it has his own foundation now. So, depending on the future of AIR, if Adobe decides to stop the investment... we will have to move forward.
Haxe is an amazing language, with a really helpful community, it has a learning curve for advance topics like Macros, and sadly, very poor IDE support and the frameworks (like OpenFL) are mostly way more unstable than Adobe's Flash/AIR from my experience.