Peer reviewed science does have more value than an opinion.
The purpose of publishing in peer reviewed journals is two fold. First, scientists need to communicate their findings to other scientists. Second, peer review should catch any major errors in the paper, be it methodological errors or misinterpretation of results.
Very rarely do peer reviewers repeat the experiments in the paper, nor does peer review guarantee that the results are repeatable. The whole point is that you can’t know if results are repeatable if no one knows what experiments you have done or the data that the experiments produced. The primary purpose of peer review is communication of results so that others can determine if your findings are repeatable and reliable.
Science is absolutely susceptible to those things. This is why there is oversight and peer review. Most importantly, career advancement is based on reputation. If you are caught fudging data or cheating then you could seriously damage your career. Even honest mistakes can sometimes damage careers.