Looking at the data, they seem to compare what effective[1] founders do more frequently than ineffective founders, so they do compare the two. The study sort of

Looking at the data, they seem to compare what effective[1] founders do more frequently than ineffective founders, so they do compare the two. The study sort of includes these % gaps in practices of most and least effective founders. [1] effective here is subjective and reported. They have a discussion of why they don't use other metrics like company valuations and it boils down to that data is harder to collect I think