Correlation finding is still an incredibly important first step towards figuring out what would be a good double blind randomized trial. Science media could definitely report correlations better.