You don't have to pay for an order of magnitude more people: education is not a technology yet, it could give no guarantees on a results. To get one bright student you need to teach ten people through primary and secondary school.
Harvard, Yale, University of Texas, Stanford, Princeton, Massachusetts Institute of Technology, University of Pennsylvania, Texas A&M, University of Michigan, Northwestern University, Columbia University all have 10-30 billion EACH.
Roughly 100 colleges in the United States have endowments exceeding 1 billion dollars. Even with a billion in a broad market index fund they could burn 30 million a year and likely never run out of money, I think the most expensive tuition in the country is like 60k or some such, so the poorest of those 100 schools could take at least 500 students a year for free and not go broke.
There are about a dozen universities in the United States that have research budgets alone exceeding 1 billion dollars a year.
Kinda makes you wonder why college tuition is so expensive at those 100~ schools...
And I don't think those endowments include real estate value or income from IP the schools own thanks to research conducted there.
Funding was shifted to fees. Although considering how much of student loans is not expected to be repaid it's arguable that a lot is still state funded...
You could make the exact reverse point, that universities in the US have built their expertise over decades if not centuries and Foreign students are largely benefiting more from them than whatever they could get in their home countries.
Except that this isn't true at all.
And it's really, really actively fighting against it: electing Trump (and the consequences with it) is not the least they did in that perspective.
This means that as a country you don't have to pay for the primary and secondary schooling of all these people, but you get to reap the rewards of them working in your country and contributing to your scientific lead, to your economy.
But like all advantages, if you don't pay attention, you can squander it. You can make it easy for smart Chinese students (in this instance) to do science in the US and benefit US academia and businesses, or you can make it hard, and the same people will instead benefit some other economy.
It's not about diversity, it's about not realizing that the free firehose of cheap and willing smarts and labour isn't perpetual, that it actually has to be maintained and encouraged.