Might be lacking a lot of details, but I don't think I could trust an open source plane. But I would trust a proven kit design.
Hmm, this is actually a topic I feel almost qualified speaking to.
I don't think that with current materials we'll be printing planes out of 3D printers. I could see printing odd parts though (non-structural plastic pieces of molding/etc), and that could save money. Plus most of these are parts where safety isn't as much of a factor (there will be exceptions).
Open source and quality are completely different things. You can have either, neither, or both. For something like an aircraft I would imagine that people would collaborate to get the design to a point where it is ready to prototype. The design would be versioned at that point so that the prototype could be tested. That's basically how any manufacturer would do it, and nothing about the quality process prevents it from being open source.
The bigger issue is going to be financing the work. Open source works great for stuff like computer software where testing isn't critical, or where it can be automated. When you're talking about an aircraft then the only way to test it is to make the thing and spend many hours operating it or subjecting it to test procedures (wind tunnels, flex/stress, etc). That sort of work is very labor-intensive, and flight testing involves insurance, fuel, maintenance, etc.
That's why you don't see open source applied to things like drugs or cars or planes already - they all involve lots of testing labor that isn't interesting but which is quite expensive.
I think the first place we'll see open source applied is avionics. If you start with TSO'd sensors (GPS, airdata, engine sensors, etc), you could do all the data integration and display yourself if you have the time, and you could conceivably do quite a bit of testing automatically or in a distributed fashion (you don't have to actually have a certified GPS to simulate the output of one). If they don't already exist this would create demand for cheap TSO'd sensors that are just sensors - they don't need displays/software/databases/etc. I could see a volunteer-based organization possibly achieving certification for such a design. I think it still would be a huge challenge and not something we'll see (certified) anytime soon, but it is a way easier problem to tackle than building planes.
