The problem guys is not the final segment as that has been taken into consideration by the procedure makers wih EASA rules. The problem is the way you get to the final segment. Flying for a certain Red Nosed operator to the most northern parts of Europe it is crucial with cold temperature compensations. So even if the final segment is protected down to the temperature that is published on the chart the way from the initial fix via the intermediate segment is not protected and thus the operator must ensure terrain clearance.
Since the 737 in our case has no way to intercept a glide path from a corrected intermediate segment it will start a descent down to the final approach fix (point) and thus penetrating the terrain clearance. That is why we temp correct all altitudes below temp corrected MSA. What people unfamiliar with flying in cold weather tend to forget is that the true height of the aircraft when flying temp. Corrcted altitudes are the same as in ISA STD atmosphere and actually more correct.
As I saw someone saying they add 10% to all altitudes can be conservative or not enough. 10% is only safe to -10 degress anything bellw that has to be corrected more. And then there is of course the effect wind can have on minimun altitudes. I suggest looking at RNAV X in ENEV and start playing with wind and temperature and you will quickly find the need to correct, otherwise cumulus granitus might appear.