Evaluators are now saying the OSM introduced a completely alien workflow, produced shoddy answer-script scans and recorded marks incorrectly.