This. It seems like new features are developed, but it takes decades for the good ones to rise to the top and end up being implemented. And the academic model, of applications for grant funding works quite well for just these sort of advancements, at least in other fields. Somebody who has the money and time to put together a grant program targeting improvements in both programming languages and, perhaps more importantly, real world user experiments in the design of programming languages, could really make a difference. There is no inherent reason that that the decades for useful feature implementation couldn't be reduced significantly, given a moderate source of funding, a good journal community, and better research methodologies.