Risk threshold for alcohol consumption: 100 grams per week

Notice also that the click bait article does not say the peer reviews become less reliable. The research quoted asked that the editors judged review quality, not reliability, so not to worry there. The primary dimension of quality appears to have been based on the number of errors found in the manuscript. (If I were an editor I too would find reviews that identified more errors more valuable. It would given me something to tell the authors when I bounced the paper or insisted it had to be revised in a certain way. )

As for fraudulent, lets draw a distinction between fraudulent and methodologically flawed. In the former there is foreknowledge with potential or actual criminal intent, in the latter there is either a lack of knowledge of or a lack of understanding of how to deal with the flaws. Empirically speaking there's a very low likelihood of fraudulent work outside of domains that receive funding from private corporations and even there it is not so high as some folks like to think. As for methodological flaws, you are 100% correct that is is very difficulty for people outside the research domain to fully understand what is happening without a lot of work. (But then that's a lot like understanding the Russian language. No way to fully understand that without a lot of work.)

As far as people declaring BS on research findings, that has been the case for longer than you or I have been alive. Most people think of research as being light switch type studies and don't realize that many kinds of research are more like a maze. Similarly some people often don't want to hear results that make them feel uncomfortable (e.g., frequently drinking 5 or more beers in an evening can damage your health, drinking 3 high ABV beers a night, every night for weeks can damage your health in the long term). So they declare BS, while ignoring the fact that in the long haul, empirical research in many areas, flawed though it may be, has provided many solid answers in many areas of study.

e.g., the Brooklyn Bridge still stands despite it's age.
e.g., immunotherapy drugs are shaping up to be game changers in extending the life of certain cancer patients.
e.g., the science that underlies and supports the art of brewing has brought us a long way in terms of what is possible in terms of variety, consistency and various measures of quality.

 

100 beers a week? Cool.
Then I'm good to go.

 

This is a problem and you might be glad to know that there is a growing movement to make scientific journals open source, so that one does not need a university affiliation or an expensive subscription to access them.

Also, I'd just like to second @drtth's comment that fraud is a very small problem and what I'm guessing you were getting at was methodologically unsound studies or studies that draw conclusions which over reach what the results have indicated. It's worth noting that within a given discipline, a single study never carries much weight. If anything, science and research training tends to make skepticism the very first reaction to reading other people's findings

Your point that the way science is communicated to the public could be substantially improved is certainly correct though. Unfortunately the solutions to that problem are complicated by the fact that scientific literacy and critical thinking are not currently valued in this country, and the financial incentives for much of the media don't encourage long form or in depth reporting on complex issues.