bookworm - thanks for that - I feel suitable chastised!
Again, not my intention, but I guess I only get to claim that a couple of times.
I was not saying that it follows that one is in contravention of Articles 63/64 just because one is flying a non published instrument approach. I just think that maybe some pilots need to bear in mind that, generally, such a practice means they may not be enjoying the usual level of safety compared to an approved procedure with full ATC.
I agree that a very careful risk assessment is required.
The risk of collision with other aircraft is probably not very significant. As we've already discussed, uncontrolled IFR outside controlled airspace is standard practice in the UK. Flying over a VOR at a rounded cruise altitude like 2000 ft probably puts you in more danger of encountering another aircraft than when flying an approach in the last 1000 ft down to the surface.
Published IAPs are regularly flown without ATC, but with self-announcement, in countries such as France. The system seems to work well, and the fact that aircraft tell each other what they are doing is important. One of my concerns about the illegality debate is that making DIY IAPs illegal discourages that communication.
The risk of collision with obstacles and terrain on such home-made approaches is surely the dominant risk. There are, unfortunately, too many examples of pilots getting it wrong, of which the most prominent example is probably the crash of Graham Hill's Aztec attempting to find Elstree in 1975.
Published IAPs are produced in accordance with strict specifications that most pilots are unaware of, and they use the information from a survey to assure obstacle clearance. Without an understanding of either of those, the pilot may be in a difficult position to create a safe procedure. There may be an argument however, that it is possible to fly safely with an obstacle clearance less than the arbitrary 1000 ft but rather more than the 250 ft used for non-precision approaches in published procedures.
The flip-side of regulation is usually in the alternative risk-bearing activities that it might promote. If it were made illegal to carry out a carefully assessed DIY IAP, it would leave the law-abiding pilot with the consideration of prolonged scud-running in marginal conditions as the legal alternative to a much safer high altitude cruise and short cloudbreak at destination. It is right and proper to require pilots, via Articles 63 and 64, to exercise appropriate risk-management in every aspect of their flying. The more you narrow the options by regulation, the more difficult you make that job.