It may be useful to separate science itself (with its self-correcting nature and such) from academia and the community/economics of scientists and research. Furthermore, if the incentives for publishing research are wrongly aligned, can you really trust published papers in mass?

Scientific consensus can be a hard thing to define sometimes. Climate change is an example where there are enough studies out there that a meta-study can prove consensus, but what about something more recent and difficult to measure, like the effectiveness of covid vaccines/measures?