I can't entirely disagree with you, but no one has to run a gauntlet to publish. You have to run the gauntlet if you want to publish in a gauntlet approved publication.
There will always be occasional revolutionary ideas skipped over (temporarily), but by and large science is incremental rather than revolutionary. Good data must be explained by any theory, and any new theory must explain all the data, plus suggesting new and fruitful lines of research.
Einstein was not dealt an easy hand. The Nazis tried to destroy him. "One Hundred Scientists Against Einstein" was the name of an official pamphlet. Such things can slow progress but not stop it.
Seems to me that some of the savings from streamlining the administration of grants could be used by the Feds to collect and publish articles which were previously rejected for content.