Nodding in violent agreement with these insights.
In particular #3. The intention is to identify and quantify impact… assigning a quantum of OP to each project only nominally does this and actually makes the issues described in #1 a lot more prominnent.
If understanding what is most impactful is the goal, why not just focus on having people answer that question. It is an over-simplification but simply assigning whether a project is highest impact, high impact, medium or low impact would avoid a lot of the complexity and (hopefully) make it a lot easier to see outliers on both the project and badgeholder sides.