An observational study is a type of study. It carries value and can be well designed and rigorous. Just like a RCT can be poorly designed and lack rigor. EAch type of research is intended to assess a specific reserach question. Observational studies are valuable for assessing trends and population data, something a RCT can't do (by nature it aims to limit variables when it is impossible to limit any interaction to a single causative effect).

I agree that a low-carb/high (healthy) fat diet can be very healthy, but it can easily be botched as well. With respect to meat, the type matters. Fish, deli meat, and prime rib are not equal with respect to health.

Another consideration is goals of the diet. This article and study refer to big picture health (mortality). If you want to build muscle, you need to eat a decent amount of protein (but not as much as many people advocate for). If you want to lose weight and manage blood sugar , low carb is fantastic. If you want to promot autophagy and longevity, fasting is the way to go.

Nutrition is complex and when done well, observational studies can be a helpful piece of the puzzle.