Computer science is leading the way. Let's abandon journals, move everything to preprints, and evolve from there with some sort QC system that picks the good/interesting preprints (computer science does this via conferences).
Journals are dead. Every field need to abandon them. the peer-review process was always about selecting articles, not verifying them. Computer science has shown the way: open-prepublication (arxiv), allow the whole field to review and discuss the pre-published papers, then conferences select the most interesting one for live presentation and discussion with the peers.