It doesnt work like that. You have a proposal that you send. A group of people (never the same ususally) come and review your work with half of the people that

It doesnt work like that. You have a proposal that you send. A group of people (never the same ususally) come and review your work with half of the people that have not heard from you. They look at your story, and your metrics (papers, patents...) and put a bunch of scores (such as societal interest, innovation, approach, personel...) then the jury convene discuss a bit and change their scores if they want. The head of the jury then takes. all of that, combine and make a report. So in practice I've seen big abuses at every stage. Friends that help other friends get grants. Enemies that destroy theor enemies grants. Thieves that destroy the grant and then develop the work themselves after (that one we reported two times for two events to the funding agency that just removed that person from jury panels, nothing else). In the end, no one cares if they failed to achieve what they promised because nobody sees the grant (not public unless you submit a FOIA request) or remembers it 5 years later.