I would add to what Jeff said that there’s a significant bottleneck that prevents releasing all of the things in perfect shape at the same time: a lot of the same people do the critical work on Julia itself as well as the packages that people are miffed about the up-to-dateness of. If we had human cloning technology, we would have used it on Keno by now 
. Ditto for Jeff, Jameson and many other spectacular folks who already do more than seems possible for a single person. To improve this situation, the best approach is to increase the number of people who can work on Julia and important packages. If you’re not an industrial foundation that can directly fund development like the Sloan and Moore foundations, then the best thing you can do to reduce the bottleneck is likely to become one of those people yourself.
Since we cannot implement all of the things at once, the strategy that we’re taking is to stabilize the core language first, then turn our attentions to important packages, tooling, and compiler improvements that we can make without breaking things for people. Since the core language will be stable, we will have a solid foundation to build on. If we had unlimited resources, sure, we would release Julia 1.0 and all of the packages and tooling one could possibly want in perfect shape. But our resources are far from unlimited, so we’re taking the same strategy that Rust, Go, Clojure, Swift, and every other new programming language I can think of has taken: stabilize the core language, then once that’s done, turn your attention to important libraries and tools. That’s exactly what the Rust “Libz Blitz” linked to here is about. Rust is just a couple of years ahead of us: they released their 1.0 version a couple years ago; since then, they’ve spent a few years working on tooling; now they’re turning their attention to the package and library ecosystem.