While schema.org exists, it is a massive specification with a lot of cases that are likely unused if you were to optimize it for application use. This project really just points at the tension between using such an extensive schema versus more "practical" and application-focused ones, especially when token costs are in the question.