I don't think you understood the article at all. For example, they talk about how standard grading practices often misrepresent mastery, by including grades from the beginning of a class, before the learning has happened. This benefits people who start out with advantages, and clearly disadvantages people who end up with the same level of mastery, but start at a place where they need to learn more. This relates back to software design: Averaging the homework scores from the entire class is a "default" in how most grading software works. It doesn't have to be.