Points (1) and (3) appear to contradict each-other, AFAICT. Surely if you made researchers "waste" less time on the grant process, that would reduce the public's ability to supervise and intervene on what get's funded in an informed way? Much to most of the grant procedures researchers have to follow consist precisely in generating metrics, documentation, and other material meant to demonstrate to a skeptical public where the money is going.

You can trust professionals to do their jobs as they see fit and write them a check, or you can make them "waste" time proving to you they're doing the job you want them to do. You can't have low-trust and low-effort grant administration.