the 'holy grail' of 3d printing is something with high resolution, a toolchanger, and dissolvable support material
because even injection-molded parts need some hand assembly, but if you can print all of the parts inside your thing, you can get way better scale and efficiency than injection molding in terms of human toil.
but my (cursory) understanding is that things approaching that resolution won't reach the massive innovation-explosion for another 10 years, because, well, they're under patent.
this also held back 3d printing until the 2010s, when patents expired and the current innovation surge you see now happened.
right now resolution is bad enough that things have to be scaled up by an order of magnitude, and still require some finishing -- but we've gotten there in a decade. Tool changers already exist, dissolvable supports already exist, we're just waiting on material science or motor technology breakthroughs to increase the resolution enough that you don't need a finishing pass that can't be fully-automated. Hybrid resin-filament printers are another way that some of this can be done. But once we can get that, at high enough resolution to do injection-molded quality and cost, local manufacturing-on-demand becomes viable and you upset the entire industry, even if a first-generation machine of that type costs, say, 50-150k as a guess of what the bleeding edge would be. Because at that point, you're able to cut out almost the entire supply chain (bulk filament takes up less room on the boat than assembled goods, even!)
or, you can do my terrible idea, use semi-transparent filament... and gamma-knife the thing. multiple UV lasers (or just pure heat) that only hit the material-setting threshhold at the spot you choose. Do I think this is a good idea? no. Do I think it would allow for as-yet-untested forms of manufacture? absolutely. will it work? almost certainly not