[robbedpeter]

Scientific credibility is called reproducibility. If it's not reproducible, it's not credible, and that's got nothing to do with individual scientists, or the quality of their previous work. Peer review, as she is practiced, is an unnecessary social affectation and an increasingly onerous barrier to progress.

Journals are rent extractors targeting the "science news" demographic. Curation is the justification for their involvement, but in practice it amounts to politicized gatekeeping for profit.

Real science communication looks like scihub - the firehose torrent in which the curation is done by individuals seeking information separate from any "proper" authority.