Some of that is given back by having a lot more university-supported Postdocs, where as public R1s largely require their faculty to fund postdocs of their own grants.
It sounds like you may accounting a little differently than what occurred at my institution. The 40% was taken for administration and overhead; all faculty and student pay, hardware and software, and any other research materials all came out of the remaining 60%. Part of why the ARL balked was because THEY were the administrators and approval had to go through THEM, not the university. There's no reason the university should have taken so much for "administration" they weren't providing.
Shuttle had two loss of crew accidents in 135 flights. And no extra mission failures.
That's very misleading. After the 1986 Challenger explosion, one of the intended goals of the shuttle, to deploy and maintain spy satellites and equipment, was considered too risky. As a consequence many of these missions were shifted to other launch platforms such as the Delta rocket family. I'd argue that all of these should be considered mission failures, from the shuttle's perspective.
Even if you get a $500,000 grant, anywhere up to 2/3 of that goes immediately to the university you work for for overhead.
I worked on a military funded research project and the Army Research Labs contract administrators balked at the 40% (!!!) mandatory overhead costs. They felt it was exorbitant as they had their own people who oversaw nearly every aspect of the contract. The only thing the university had to provide was a space for us to work.
A meeting is an event at which the minutes are kept and the hours are lost.