Most "top" universities receive their ranking for the quality of the research they produce, not quality of undergraduate education. There is likely a decent correlation between research output and undergraduate education, which one can determine by looking at schools that have some of the highest achieving undergrads (such as MIT, Harvard, Stanford) and noting that these schools also are major research centers. Note there are most certainly exceptions to this, ie schools without a ton of research output but with great education for undergrads (I'm thinking of Reed College here, but no evidence/just my perception) or universities that don't do much with undergrads but are major research centers.
However, I'd mostly like to address the fallacy of attributing the high international rankings of American universities to the high undergraduate tuition. At Caltech, I've heard many undergrads say that all undergrad tution is about 5% of Caltech's budget, with a lot of the rest coming from various research grants. At schools with less of a research focus and more undergrads, undergrad tuition might make 30% of the budget. I can't find the blog post at the moment, but there was a discussion a few years ago about what would happen if MIT gave a full scholarship to all attending undergrads, and the conclusion was that it would hurt (but not cripple) MIT financially, but appeared possible.
Those same universities have selection criteria which predict success after college independent of the college attended. So it's not like they are masterfully teaching anything; rather, it seems very much like they are simply rebranding the best students.
To the extent that they also teach students more and better, this is actually damaging socially because it sucks secondary educational capital into a few locations. Lower end schools could do more and better if they housed many more bright and ambitious students. But if they're all going to a handful of schools, they're not improving the social networks at Random State.
Although elite universities might not "masterfully teach", the students at these universities still learn more than students at a lot of universities simply because the classes are more difficult (these are generalizations, not the rule). If classes are more difficult/cover more content but aren't taught well enough/not all content covered, then highly ambitious students just end up teaching themselves a lot of the curriculum, but they still know the material of their classes by the end. Essentially, even though the teaching might not be better, I believe an average MIT student studying X comes out more knowledgeable than student from [[ state school ]].
And its an interesting thought to distribute top students. Assuming a fixed quantity of "bright and ambitious" students, wouldn't sending more of them to lower end schools weaken the social networks formed at elite schools and destroy something unique about American education system?
However, I'd mostly like to address the fallacy of attributing the high international rankings of American universities to the high undergraduate tuition. At Caltech, I've heard many undergrads say that all undergrad tution is about 5% of Caltech's budget, with a lot of the rest coming from various research grants. At schools with less of a research focus and more undergrads, undergrad tuition might make 30% of the budget. I can't find the blog post at the moment, but there was a discussion a few years ago about what would happen if MIT gave a full scholarship to all attending undergrads, and the conclusion was that it would hurt (but not cripple) MIT financially, but appeared possible.