Back to my original point, in what way does producing high level doctoral level research lead to high quality UNDERGRADUATE education. Don’t cite the outcomes. Cite the leading indicators you believe create the effect.
Personally, for many of the schools in the list, I believe the correlation has nothing to do with causation. It’s a self perpetuating phenomenon based on list based reputation. That attracts high level students and they succeed in spite of the less than ideal educational arrangements those schools present them (giant classes with instructors given no training or incentive for quality teaching).