Always read full text of the original scientific publication before making any conclusions. About 30% of news releases of biomedical research are wrong. Imao, in neuroscience popular news/blogs this number is somewhere close to 100%.
Yes!!
For me, the key indicator is the word “might” in the title of the article, any qualifiers (“under certain conditions”), as well as being as close to the (reputable, peer-reviewed) source as possible. And the good old “extraordinary claims require extraordinary evidence”.