It is believed that unique coverage calculation and traversing the model multiple times just to calculate per-test unique coverage is serious performance hit in reporting for large projects (LOC, #classes or #tests). Give users the ability to turn this off until we find a much better way to speed this up.