I think, (being too lazy to look at the protocols myself right now) because in both cases the jumps are counted as a sequence, and the scoring for a sequence is such that the total of the two jumps included is multiplied by a facor of 0.8 (to make it not as valuable as a true combination, where the values of the individual jumps would be added togather, no factor applied). However, the amount of points from the single jumps on the back ends of these sequences is so negligiable (or maybe even get no points at all in a combination?) that having them added does not overcome the reduction on the points for the first jump, when the 0.8 factor is applied. ... A bit of a silly/debatable way of scoring sequences, but there you go (if my understanding is correct)