2) Very few schools have endowments that are large enough to support current faculty research costs; even Harvard can only support all research off their endowment for about a year.
3) Endowments are now taxed, so they will have even less available for research.
2) Endowments aren’t the only way to fund research, and not all research is equally valuable. Some is probably negative value, given the replication crisis.
3) Investment income from the endowment is taxed at a maximum of 8%. If that’s enough to break the US university research machine, I’m not sure it was ever working in the first place.
The ROI is not positive because there is no guarantee basic science will lead to any money-making outcome. That's not how basic science works and it is never how it worked, but basic science is absolutely crucial for advancing tech. If you can convince shareholders that it is e.g. worth investigating unusual crystal structures with no intended product in mind then people will gladly share some grants with you to edit so you can work some of that magic. It's just not feasible, no company would pay for that, but sometimes it leads to important discoveries that change the game.
And I only mentioned endowments because the parent of my comment did, but again, the important point is that endowments are not intended to entirely fund the research machine, and never will.
I’m open to the idea that government can play a role here, but only in a very small way. The government is spending hundreds of billions of dollars on research. They are such a huge consumer of talent and resources that they are crowding out private initiative. This ain’t Bell Labs of the 1950’s. And if you asked those researchers what they think of today’s system, I’m sure they’d be appalled.
2) Harvard has a $54 billion endowment and spends $6.4 billion a year, that doesn't include tuition and fees, that's just what they have in the bank.
3) Actually spending the money is tax deductible and better for the economy.
See my other comment on the likelihood that any corporation would want to pay for basic science. When would any company choose to fund something that would not guarantee a return on investment? It goes against the nature of the goals of a corporation. I would love to know of for-profit examples of this. I'm sure there are a few, but I doubt there are many.