What I have done is made the score calculation work by
1) add up all the points the user got for each item
2) adding up all the points for every item answered in a session
3) divide the users points by the total points
4) multiple by 100
This takes into account assessments where all questions do not bear points. It does not allow variable weighting of sections etc. This feature appears unfinished.
There are two columns for section points 1) as_sections.points 2) as_assessment_section_map.points.
It's also not clear variable weighting is useful for general users of assessment and it could add great complexity to the user interface. I can imagine statistical analysis of the points per section might be useful, but for a standard multiple choice test, its overkill.
