Your example is perfect. The 25-5 team is likely better, efficiency-wise, than the 30-0 team.
One of my points here is that it's not untenable to have KU ahead of Baylor. Would I do it? Probably not. I'd probably rank Baylor ahead of KU. But based on the data available, it seems to be telling us that KU is slightly better, efficiency-wise, than Baylor, and that makes a decent case for ranking KU ahead of Baylor.