I wish there was a sort of Snopes/Politifact/Rotten Tomatoes for science - an independent body that queried the top scientists in each field, and used this to quantitatively map the current state of the art and the consensus (or lack thereof) around that field's theories. Instead of e.g. 'the current state of CRISPR research' or 'P vs NP' being something that lived implicitly in the minds of the few researchers that keep up to date with all the relevant papers, it could be explicit on the site for all to see (including journalists, who would finally have an easy way to avoid putting their foot in their mouth with another 'Scientists show that lung cancer causes smoking!' headline). Even if all that could be said in a given area was 'Correlation has been shown, causation has not, a study is underway to try to establish this, and 60% of randomly queried researchers in the area believe causation will be proven' then that would still be super valuable. The result "99% of climate researchers and meteorologists believe climate change to be manmade" is a great result, but it would be much more hard-hitting if it came from an institution that was already considered to be reliable.