I remembered reading a translation in another forum of a Lakernik interview from a December 2017 issue of the "World of Figure Skating" magazine that has an interesting point that I'm quoting here:
Interesting... on the one hand, December 2017 is a long time ago. On the other hand, it almost sounds like the positive GOEs might work the way I want them to. As long as positive GOEs can’t offset negatives to more than +3 GOE (or can’t go positive at all), then the risk-reward isn’t nearly as dire as initially proposed. Which is good news for women’s singles, arguably not so good news for men’s... this is all such a mess. (I still think the quad rep proposal has more potential consequences for the men but that’s a question of it passing.)
Do you have a link for a document with this? All I've got is the USFSA's guidelines
http://www.usfsa.org/content/2017-18 SP GOE Guidelines.pdf which just has the 8 bullet points including good height and distance. However I've heard the Eurosport commentators (Mark Hanretty) talk about extra credit for example for 2nd jumps in combinations being as big as first ones and he is very accurate when it comes to these sort of things. If you or someone else has a further breakdown of the guidelines I'd be very interested.
From what I understand, bullets 1 and 2 (entry) are applied to the first jump, and bullet 5 (exit) applies to the landing of the final jump. The rest apply to all jumps in the combo, with 3 (varied air position / delay in rotation) being an oddity and one that can cancel out another positive feature (if you go for a tano but in doing so lose bullet 4, you’ve created the same number of positive features); you also don’t need to hit bullet 3 for all jumps in order to get +3 GOE for the combo, nor do all jumps in the combo need to meet that particular feature in order for a skater to be awarded credit for it.
It’s not that there’s “extra credit” so much as that a quality -2T/-2Lo should be worth more to the combination’s total GOE than a small, laboured one. A skater shouldn’t technically be awarded bullet 4 (great distance and height) if the second jump in the combo is lacking. Even if you still wanted to award the first jump bullet 4, however, you could theoretically still hold the combo under +3 GOE (as you should) by not awarding bullet 7 (effortless throughout).
So, yeah, each jump isn’t awarded individual GOEs so much as that entry and exit GOEs are applied to the first and final jump, respectively, while the bullets assessing jump quality are relevant for each jump in the combination. That’s my understanding, at least. [emoji3]