Both J and Eric are right.

If some tyro builder (me for instance) had ready access to a CNC mill, then plans start to look more appealing as opposed to kits. For instance I can hand build a couple representative ribs, then get the mill to make the rest, all exactly right, in far less time than I'd spend screwing them up. I'd only need to pay for (some portion of) the mill, the raw parts, and cleanup. OTOH, I'm never going to financially compete with some factory's economies of scale. Nor expertise.

------

The "open source" part of this is ready availability of plans for well-known, proven, approved, parts that we could make ourselves rather than purchase. As we build up a body of items and parts useful for all airplanes, then more and more opportunities arise for totally new and/or custom aircraft design. It's not so much about making a complete OS airplane ... yet.

For example, suppose we make good wire harnesses, axles, struts, ribs, antenna ground planes, engine mounts, control horns, etc. with CNC and make the plans available? Once proven and approved, that makes the FAA examiner's job easier too. All kinds of things we can now reliably make ourselves.

While not everything can become open and/or automated, we don't have to stop with hardware parts. How about an avionics stack built on reliable solid-state sensor and an android tablet? Or software-defined radio? We probably can't do open-source looms for carbon fiber, but our local club might be able to do some kind community vacuum bagging tool.

3-d CAD? I will look into it a bit. I have not played with any of it so don't know much. Google has something free. Autodesk just released a crippled free version of Inventor. (It will do components but not assemblies.) No idea what SolidWorks might or might not have, nor what's on FOSS software places.

The second big starting issue is "repository." Do we know where and/or how we want to store and distribute the designs?