I mean, it kind of depends on your goals and how many bees you are willing to lose. A novel pest like this will take out colonies no matter what, so I'd prepare myself for the fact that some losses will be inevitable.
There is an extreme end of the spectrum that says to just live and let die in a situation like this. Let colonies get infected and cull/requeen or simply let varroa take out the weak ones. It's natural selection, the strong survive, and the colonies that are left will be more resilient. The downside, of course, is you could lose everything. The other extreme end of the spectrum is to always treat, on a schedule or whenever colonies hit a predetermined threshold. The downside here is you are breeding bees that will always need to rely on the treatment.
Neither of these options are wrong, but I personally prefer a more middle of the road approach. Basically I'm at the point where I'm willing to have to treat with oxalic acid. It's cheap and easy to apply, so for the moment at least, I'm treating all the colonies with at least 1 round of OAV in the winter, as a little bit of an assist. If I have a colony the needs more help than that, they get requeened. I will treat a crashing colony with something powerful, like FormicPro, to keep a mite bomb from spreading, but that's all.
I'm also kind of off of the idea of treating based on a % infestation alone, because some strains of bees are resilient to mites, as opposed to resistant, meaning they can handle a high mite load without crashing. So I'm more inclined to eliminate a colony from reproduction based on symptoms of PMS or really high mites counts, like double digits. I'm less focused on keeping numbers down, and more on finding out which colonies are able to handle themselves on their own, because my goal is to use as little treatments as possible and someday, hopefully none.