The point I was trying to make about approved vs home built approaches (and, to an extent as well, cloud break procedures) is that you get a much better indication that you are deviating from what you KNOW to be a safe profile on an approved IAP than on what you THINK is a safe profile on a home built approach.
How do you get "a much better indication"?
My own view is that, purely from a flyability standpoint, there is a world of difference between descending over the sea and then following a home built approach that you have programmed and stored into your NAV system and flown several times in VMC to test it to an airport with no hazardous terrain vs making up a descent and approach into a mountain airport when you get stuck in unexpected IMC.
I agree with that, and flyability is an issue. However, there's nothing difficult about setting up a safe trajectory into Egelsbach. You don't approach it down a valley. If you simply put in waypoints and calculate minimum levels to establish a 3 degree descent (preferably 4 degrees, but 3 degrees doesn't hit anything) then you fly a safe trajectory. That the Citation crew did
not do that contributed to their demise. It would be possible to put an LPV into 27 at EDFE as far as the final approach segment goes: the problem is the missed approach segment and the conflict with 18 departures from EDDF, which is why there is no published approach.