> Otherwise, any source I provide will be met with "yeah, sure, scientist Y wrote so in journal X, but journal X has been wrong before, so your source is wrong."
I think it's easy to gloss over the fact that science exists within a context, and if you aren't a participating part of a scientific field, you probably won't have full access to that context. This is a leading factor behind how in many debates, both sides are waving around scientific articles.
This has been especially visible in the last few years with the pandemic and the debate around stuff like restrictions and masks and what have you. It takes years to produce quality science, and here we were months into a pandemic and people kept batting each other over the head with preliminary scientific findings as though Moses himself walked down with them from mount Sinai. The truth was we didn't have all the facts, the scientists for the most part were pretty clear about this, but none of that seemed to matter to the armchair epidemiologists engaged in fierce internet-debate.
If p=0.05 is sufficient confidence to publish, one in twenty published findings will be wrong. In one study, that isn't bad, but in tens of thousands, that's a lot of dodgy results. In practice, it's probably far worse than that because of publication bias: Scientists tend to want to break new ground and publish extraordinary findings and maybe won't publish boring stuff that just reconfirms that energy is still conserved or whatever.
Many just google up some abstracts off arxiv that seem to support their claims and toss them out like pokemon. That's honestly only marginally better than basing your arguments on proverbs, events from the Iliad, and or bible verses.
Science just doesn't work like that. Individual articles are not infallible doctrine. Taken as a whole and in context they do form a pretty compelling argument, but individual articles on their own, even published in major journals, are absolutely fallible.