I agree with most of the author's points. But trying to publish negative results is hard and can hurt your career as they don't get many citations. What he is proposing will make us better researchers but will kill our careers to the point we aren't researchers, defeating the point.
To prevent the circular reasoning, we need journals to enforce some rules or have higher standards. And I don't know how why they would do that. Which is rather unfortunate and make you wonder what has academia become :(