Protection all the way to the runway TDZ?
It is a hot potato specifically because both TERPS and PANS-Ops have largely ignored this issue of assuring a safe flight path below DA(H) to the TDZ since the concept of original TERPS and before before CRM was invented (and well before 1975). Hence ONLY [real] RNP protects all the way to the TDZ, and all the way back out, for a balked landing from the TDZ, considering both rare normal and selected non-normal events (e.g., the first nav failure, and the first engine failure on a multi-engine aircraft). Hence NO TERPS or PANS-Ops procedure would ever pass a serious scientifically derived SSA, FMEA, or FHA, at anything other than a largely trivially un-obstructed runway. This was one of the original and continued flaws in both TERPS and PANS-Ops, which led to RNP, and for which (other than the well configured and signed AC120-29A), still has unnecessary, counterproductive, and flawed criteria even being used by FAA in their other RNP references for their version of pseudo RNP [lite] (e.g., AC90-101A and AC90-105). This is why airlines using [real] RNP procedures typically use criteria related back to the criteria of AC120-29A, and not other more recent FAA RNP related criteria.