A few years back there was a big hubbub around flossing from an AP report that mentioned the lack of gold standard randomized clinical trial (RCT) evidence for flossing. It caused a lot of takes that flossing was unnecessary and a big lie. But really it just exposed the challenges and limitations of RCTs, and why we shouldn't take the absence of such evidence as evidence of ineffective treatment, ignoring the cumulative body of evidence and theory.
This has been a long standing debate in the development economics field as well. With the revolution in experimental economics, RCT or nothing has become like a religion in the space. And in part for good reason, as a well-designed RCT can tease out if the effect is truly causal or not. There's been a number of significant reversals based on RCTs.
There isn't much gold standard evidence for flossing. But there's a lot of lower quality evidence and clinical experience supporting it. And the mechanistic theory is as straightforward as germ theory of disease- food particles lead to gum disease and cavities and flossing removes this debris.
In the same way there's mixed RCT evidence for masks with the coronavirus, but a lot of other evidence besides RCTs to support masking. And the mechanistic studies are all clear on masks as well that masks filter viral particles with infection and disease dependent on viral exposure.
But in both cases, the proponents of RCTs don't acknowledge the limitations of such trials for the interventions. RCTs face ethical hurdles. If you know a treatment likely works it can be difficult to justify withholding it from the control arm for a long period. Furthermore it is impossible to blind masking and flossing for the participants. Likewise monitoring is a challenge. You can't ensure that the control groups aren't flossing or masking, and you can't ensure the treatment groups are actually doing them either. So you may be underestimating the treatment effect if you have mixed adherence. This was seen in the Bangladesh and Denmark mask RCTs, where mask wearing was fairly low in the treatment arms.
This isn't to say that RCTs are bad or unnecessary. But at times there can be a cultish focus on RCTs without regard for all the other contextual body of evidence and knowledge that inform the question. Most of all the absence of such evidence doesn't necessarily mean there is evidence of absence. And one always has to be conscious that RCTs can have their own weaknesses and may not be answering the questions you think they are.