I see a lot has been written since I started this post, with interruptions, so I need to read them. However, I want to point out that values of GOE go both ways. The flip side of proportionately lower penalties on low quality is lower rewards for higher quality. If GOE value for a quad is increased for the purpose of penalty, it will also increase the difference between a front runner with quads of high quality from those attempting to close the gap. An unintended consequence?
Not necessarily. A couple of years ago the way the GOE values were set for quads and triple axels was in increments of 1 point (1, 2, 3) for positive GOES, same as for triples, but in increments of 1.5 (1.5, 3.0, 4.5) for negative GOEs.
To me, that was the best approach and I wish they would go back to it. Apparently quad proponents were successfully able to argue against it, however.
And at the same time the negative GOE values for triples have also been reduced, although the positive ones are still in 1-point increments.
For lower-value elements, it had already been the case that the negative GOE values were smaller than the positive ones.
The precedent already exists that the values of positive and negative GOEs for the same element are not the same.
If anything, I'd like to see the positive GOE values raised for some elements to encourage quality, or nonquantifiable difficulty of execution, by making +2 or +3 on an easier element worth more than base value on a more difficult element. (How much more difficult to be negotiated)