I wonder what criteria a pilot would use to design a DIY approach? Many years ago I spent several weeks on a Instrument Approach Procedure Design Course. If only we'd realised you could just cuff it with a GPS and the back of a fag packet we wouldn't have had to learn about navaid accuracy, obstacle clearance sectors, acceptable climb and descent rates, weather minimas, missed approach segments etc etc... if only (it was one hell of a course!). Seriously... DIY is for bodgers!