The problems you describe are an inevitable result of capitalism. You touched briefly on the incentives of publishers by referring to a certain cohort as “status-quo partisan” publishers but, in fact, publishers, editors, and scientists all have conflicts of interest with regard to the truth vis-a-vis what monied interests want the truth to be.
The entire journal system is screwed up beyond repair on account of this, and the “publish or perish” paradigm at universities rests on a compromised system of science journaling. That p-hacking occurs at all tells you all you need to know. The only way to overcome this is to take vested interests out of the equation, and stop lending academic credence to a clearly outmoded method of scientific journaling.
The way to address this is to stop publishing any results as credible science until after they have been replicated, and all failed attempts to replicate a study would also need to be included in the publication. Someone who completes a study could solicit in a public forum for others to attempt replication, and the entire process could be documented. It could then fall to the community to determine why a replication attempt failed, or if a successful replication is significant, instead of to partisan editors to serve as gatekeepers on these matters.
All that said, I think it is highly unlikely the scientific community will ever reach this level of credibility and ethical uprightness within a capitalist context. It will have to start with the scientists, themselves, rejecting the system as it is now and agreeing to participate in something new and better.