Why should findings from pottery generalize to reading? It's not like reading is an activity with a clearly defined goal for each attempt, where one gets clear feedback based on which to "improve".

If we were discussing a textbook where "reading" equates to solving exercises, I might agree with you. However, attempting to apply a metric-driven approach in a field where metrics are vague/implicit (which books are "high-quality", and what value does one get for the time spent reading them?) sounds like a fool's errand.