It's nit picky I know - but that does not say the judges may assign different weights to different bullet points.
...Only that they have some discretion in how many bullet points should be met to award a particular GoE.
That's more or less my understanding.
The guidelines suggest specific GOEs for 2, 4, or 6 bullet points but don't say what to do if there are 1, 3, or 5 points met. I take that to mean that if the element meets an odd number of bullet points, judges can use their discretion to decide whether to round up or down -- depending on how strongly the element meets each of those points and also on whether there were any technical weaknesses that wouldn't merit an actual reduction but would mean the element was starting on the low side of 0 before counting the bullets.
If there are significant errors to be reduced for after the positive bullet points, then it would be best to start by rounding down on the bullets.
Now, could/should a judge give +2 for an element with only two bullet points but for both of them the element was not just "good" but exceptionally good? What about
The original GOE guidelines from back ca. 2003-05 looked at each element in phases and recommended +1, +2, or +3 depending on how many phases were satisfactory, good, or very good or outstanding. I can't remember the exact wording, and it was a lot more confusing to follow than the current guidelines. But that did allow for taking into account degrees of goodness, not just an binary yes/no answer for each positive quality, so I think that sort of discretion would also still apply -- and would of course be subject to each judge's evaluation of how good the element is.
And I do think that, even if they don't do it consciously, different judges will tend to favor different bullet points. Maybe because they always remember to look for some qualities, or those qualities just jump out at them during the performance, whereas they have to consult the sheet to remember others. Some judges may be most impressed by athleticism (size and speed), others by technical control, others by artistry (attractive positions, musicality).
As long as each judge is consistent for all skaters in where they draw the line between adequate and good, I don't think it's necessary for all judges to come to the exact same conclusions on every element.
That's why we need a whole panel of judges, from which to draw consensus.