I’m probably just tired but re-reading my comment I think it’s reasonable. To clarify, the workload for all the testers could be proven to be the same while the number of things to test could have reduced by 7/8th as the article suggests.
Oh I see, you were arguing that "workload" meaning "number of things they did" was the same, not that "workload" meaning "number of things they needed to do" was the same.
