Capn Bloggs,
On particularly the first chart posted by OK465, I very much doubt that an RNP-AR approach could be designed over that ground track for "every time, IMC". It looks like it is essentially a visual downwind and close left base; wouldn't that be far too close for a "proper" RNP approach, Aterpster? I would wager that this was designed to fit in with tight airspace; the alternative, a "proper" RNP approach, would cost more and probably complicate the surrounding airspace.
Correct. Mr. Imrich knows about minimum lengths of TF to TF legs, depending upon course angle change. Make it too short and the FMS will throw up its electronic hands. And, there is a limit on making RF legs fit as well. Their respective radii cannot result in defined excessive bank angles nor can the radius be less than the RNP value being used. The RNAV visual at Ben Gurion Airport has TF legs on the turn to base and final that grossly violates RNP requirements.
RNP is indeed performance based navigation with accuracy and repeatability. But, there are defined limits on what an FMS can do to "stay on the rails" so to speak, TF or RF legs.
Finally, not all airplanes by any means, have the necessary redundancy and sophistication of FMS and avionics to perform safely at RNP values of less than 0.30.
Finally, at some locations terrain and other obstacles are the ultimate "show stopper."