A narrow product has fewer questions to answer. That is the sharpest limitation of the rule this site is built on and there is no version of it that removes the problem, so it is stated here rather than smoothed over.
The clearest case is on the register right now. One book has no online wallet at all, no casino, and settles through agents, which means seven of the eighteen questions are never put to it. It answers almost all of the eleven that remain and lands in the top band. A full sportsbook and casino has all eighteen put to it and has to answer nearly all of them to get to the same place. The two are not being held to the same test and the site prints the denominator beside every count so that a reader can see it.
A second limitation sits alongside the first. The sheet asks what a bookmaker publishes about its product. It does not ask whether a bookmaker publishes a responsible gambling page, a privacy policy or a licence statement, and at least one book in the top band publishes none of the three. That is a real gap in the questions and it means the top band, on that book, is measuring less than it looks like it measures.
A third: nothing on this site is a test of behaviour. Every answer here is about a published claim. Whether a payout actually arrives in the time an operator states is not something this desk has measured, on any book, and no page here implies otherwise. The strongest thing a disclosure grade can tell you is what an operator has committed to in writing, which is what you would be holding it to if something went wrong.
What the bands do carry, and what nothing else in this market does, is the difference between a bookmaker that will not tell you something and a bookmaker that has nothing to tell. Those two look identical in every table that reports a blank cell, and they are not the same claim about a company.