> Well that's the problem, we don't have better evidence.
We actually have had it for quite a while. The CAMBRA MOOC is put together by a dentist/statistician and lays out a point system based on correlation between patient behaviors/health and carries. It has been iteratively corrected based on effective and ineffective results when modifying behavior according to the system.
Notably in that system, floss is not given a point. Whatever effect it may have is either too weak or better described by other factors flossing correlates with.
When points like that come up, the teacher stresses that the system is only measuring odds of near term carries, so flossing may be effective for something else with regards to oral health or preventing one of the negative factors.
The sentence above was about considering the extreme case irrelevant.
Focusing on flossing itself, as person who started doing it regularly only as adult, I think that researching on how frequently one should floss (or not) is in a way like researching how frequently one should change underwear. It may make sense theoretically, but ultimately all it requires is just to pay attention.
I never noticed the bleeding and/or irritation of the gums (which dentists always mention) until I started flossing regularly.
Now I do feel a difference (as a matter of fact, I floss irregularly, depending on how the gums feel). Of course, I'm pretty sure I could survive without flossing at all.
This is in line with what dentists generally suggest. Given that flossing can be ultimately considered as simply extra cleaning, and that the presence (or not) of the effects can be easily observed, a couple of weeks of flossing should be enough for anybody to make a solid judgment more than any theory.
Well that's the problem, we don't have better evidence.