My takeaway from this is that there is no active attempt to influence the direction of Go to fit Google's agenda and that's a good thing. However, active influence isn't necessary to shape Go's direction and functionality. Probably the best example of this are the twin canker sores of GOPATH and dep. Anyone who has wanted to develop multiple projects in isolation has run into issues with these two and the way that they work-but-don't-really. For Google, this doesn't and will never matter because use of blaze / bazel is assumed and any problems relating to GOPATH or dependency management were likely addressed eons ago.