It's not about relaxing standards, but making it harder to claim causation without any further evidence of causal mechanisms. A correlation will always be nothing but a correlation, even with the most sophisticated analysis tools. Of course, this depends on the concept of cause.
What we do in CauseHealth,
@EBMplus and PhilPharm is to discuss what is causation, as such. There are a number of concepts of cause out there, and different scientific methods latch on to different concepts. We argue that none are perfect for picking out causation.
So what we write in the BMJ letter, are just some common conclusions of our research, which all have slightly different takes but work on the same types of issues: the relationship between scientific methods and the philosophy of causation (ontological and epistemological).
What CauseHealth argues, which is perhaps different from the other research projects, is that "what works" causally depends on the local context and what properties are there interacting with the causal intervention.
This is based on a particular philosophy of causation, where causation is seen as dispositional, qualitative, context-sensitive, complex and singular. We spend a lot of time comparing it to the Humean notion of regularity and difference-making theory.
CauseHealth is the 3rd research project that is developed on that basis in our work. Before looking at causation in medicine, we worked on causation in scientific methods. In CauseHealth we apply it more directly to healthcare and medicine.
I don't know about you (and I cannot see the tweets of the locked account), but some ME-patients are not happy with being told that they should try the treatment that is shown to work for many others. In our version of causal singularism, there is no one size that fits all.