I think the Beechcraft Bonanza deserves special mention here. I'm sure all the people that worked on it were experts too!
The big problem with this analogy is that it conflates three very important things:
- GA is more dangerous, period. Doesn't matter whether you build the plane yourself or if you bought it ready made (hopefully new, hopefully very well maintained if second hand)
- GA craft tend to have less experienced pilots than airliners, but even airliner pilots tend to do worse as GA pilots than when they're at work. The reason for that is simple: the processes are what keeps commercial aviation (mostly) safe.
- GA craft tend to kill the pilots, because they are more often than not the only person on the plane.
- GA craft have malfunctions like larger aircraft, there is nothing special about them in that sense. But there is something that they don't have that larger aircraft do have: redundancy. In electronics systems, in the design of the mechanical bits, and finally in the people.
- GA craft that are designed and built by their operators are experimental class for a reason: they are untested and so more likely to fail than the ones that are certified. The design processes for commercial aircraft are nothing compared to the design processes employed by what we'll call hobbyists to distinguish them.
- And finally, even though it is a fun analogy I only meant it from a skin-in-the-game point of view, a GA hobbyist is still going to do his level best to make sure that he's not going to get killed. Boeing executives only care about the bottom line, safety is a distant second. And based on my experience with the difference between the guts of various bits and pieces of avionics and the software that they run on compared to my experience looking at medical devices, their guts and the software that they run on I would be more than happy to bet that the loop hackers know as much more more about the failure modes of these devices as the manufacturers do.
Cleanroom manufacturing under sterile conditions is the main differentiator here, and that just applies to the hardware, and it is an art that the medical industry understands very well. Electronics is already at a lower level of competence and their software knowledge tends to be terrible, not to mention the QA processes on said software.
Programmers working for corporations don't necessarily suddenly grow an extra quality brain when they do their work.