I would argue that "science" is less likely to take things out of context than individuals. Individuals who read these studies out of context take them out of context. It seems to me that people not involved in a scientific field tend to read the "conclusion" of a study and interpret that to mean something like "the final say on the issue", when in reality it is just an attempt to understand the data set of the study. If you do not know the relevant literature on the topic, and do not know where whatever study you are looking at fits in with the available literature on the subject, then you are very likely to mis/over interpret the conclusions. For instance, criticizing "science" because some studies on the Mediterranean diet don't comment on the fasting of some of their population ignores the fact that "science" has in fact done the studies that suggest calorie restriction may improve one's lifespan (not just fasting...it is recognized in scientific articles that small portion size and longer meal time may contribute to the French paradox, that the cultural practice of eating to 80% fullness may contribute to the long lifespan of the [traditional] Okinawan diet/lifestyle, etc). How do you "prove" these things? It's a tough/complicated issue when dealing with open systems outside a lab.
"Science" is not taking the Mediterranean diet out of context, any more than it is taking fasting out of context. It is trying to study the variables in isolation as best it can (good luck with that...that's gonna be a while!), or just trying to accumulate statistical data to evaluate relationships between variables to develop further theory and testing. It's up to the experts to figure out how the data fits together...which means, again, that unless you are an expert in the topic familiar with the relevant research, you probably aren't going to quite understand it completely. Studying one or two studies in isolation from all the other studies on any topic is useless.
Take a study from the article above in the OP...I did a brief dive since it's been a slow night on call. The results are not unexpected...they lead to more questions to study. Part of the results: "In a subgroup of individuals of different socioeconomic status but sharing similar MDS (Mediterranean diet score), diet-related disparities were found as different intakes of antioxidants and polyphenols, fatty acids, micronutrients, dietary antioxidant capacity, dietary diversity, organic vegetables and whole grain bread consumption." In the conclusion, "These nutritional gaps may reasonably explain at least in part the socioeconomic pattern of CVD protection from the MD (Mediterranean diet)." The article referencing this study quotes an epidemiologist regarding the study, "In other words, a person from low socioeconomic status who struggles to follow a Mediterranean model is unlikely to get the same advantages of a person with higher income, despite the fact that they both similarly adhere to the same healthy diet." I think it is very important to recognize that these scientists are using phrases like "may reasonably explain", "is unlikely to get the same advantages". They are not making definitive claims but are suggesting explanations for the data, hinting at the path of future study. The way I read this, noting I am not an expert in this field, is that the diets differed in some way that is not reflected in the "MDS" (mediterranean diet score), so that there were "nutritional gaps". In short, even though both the wealthy and poor had similar "scores" regarding the "mediterranean-ness" of the diet, the devil is in the details and there was some measurable difference in the foods consumed that may explain why the poor have higher CVD (cardiovascular disease). And the differences listed in detail above suggest to me that you should probably eat your veggies and fruits, make your grains whole, and buy things as fresh and unadulterated as possible. Hardly earth shattering. But time will tell and we will see, I may be wrong...
IMHO, when it comes to general scientific answers to complex questions, the people I trust the most refuse to dumb things down, and end up answering questions like a scientist who seems to be for the most part well respected around these parts, a certain Stu McGill, who on podcasts often prefaces his answer with..."It depends."