Yes but TAing might take 30 hours a week. Anyway, MIT professors should be the best in their field, if they can’t support the small amount more it takes for the grad student with the NSF grants that sounds like a problem
MIT overhead is roughly 2/3 of the money which comes into NSF. Perhaps this might not be a problem to raise (it's generally not), but that doesn't make it okay. The bigger problem is your tax payer dollar passing through overhead into graft. Your taxes are contributing to the MIT yacht club ("MIT Sailing"), million-dollar salaries, buildings costing a significant fraction of a billion dollars (MIT Stata Center), fancy faculty clubs, etc.
Is that good use of taxpayer dollars? Of tuition? Of donor dollars? That's ultimately where all this excess comes from.