I think you're asking if we plan on recommending to developers to use HTTP2 Server Push.
We're not.
Currently, the consensus among web performance engineers and browser engineers is that h2 push rarely provides benefit.
As a rule of thumb, I'd suggest that if you were considering h2 push, you'd want about 2 full weeks of dedicated engineering time just to measure and analyze the performance. Like I said, most folks who have done this have seen their perf regress rather than improve.