I'm a physicist and only rarely get to teach math, but I suspect the situation is about the same in both fields. I've visited a fairly decent-sized sample of classes for adjunct and tenure-track review at the community college where I work. My observations from visiting these classes are of course not really what you asked for, but I suspect they're pretty representative of the US, because we're in a big metropolitan area (LA) and hire from all over the US.
The evidence supporting active learning in physics has been pretty definitive and largely unchanging AFAIK since about 1995, when I started teaching. See Von Korff for a meta-analysis. The evidence in math seems somewhat less definitive, but we've had other questions on this topic here, which you can look for.
What I generally observe when I visit classrooms is that about 3/4 of faculty do nothing but straight lecture. The other 1/4 do mostly straight lecture but also make some cursory, ineffective use of practices that are spin-offs from research-based active learning, e.g., computer polling or gamified quizzes.
My impression is that the evidence-based movement toward more effective pedagogy got going a little ca. 1995-2000 and then died out -- not because the evidence had changed but because of other factors:
(1) Faculty don't know about the evidence and have never seen evidence-based practices modeled. In math, I think the evidence is also not as definitive as in physics.
(2) The techniques take practice to do effectively. (I have also seen them modified and misapplied by touchy-feely social justice types in a way that degrades intellectual standards and disrespects the legitimate role of the written word.)
(3) Many students will complain, which can be devastating for adjuncts and non-tenured faculty.
(4) The advent of ratemyprofessor, mymathlab, and chegg have caused a breakdown in faculty's ability and willingness to realistically evaluate students' level of mastery, so bad results of traditional practices don't get perceived as bad results.
I think evidence-based teaching in physics never got going at all in K-12, and survives today at the undergrad level only in some specific places that have their own educational research programs. Examples would be UW and possibly Harvard.
I'm sure there are quite a few math teachers at the high school and community college level who spend some significant percentage of class time on techniques like think-pair-share and are very effective with them, but I don't think there's any significant number of people doing the kind of systematic, wholesale active learning approach that you may see advocated in your education classes. It would be interesting to see studies comparing the effectiveness of that kind of more modest mixed format with more aggressive active-learning techniques.
Von Korff et al., "Secondary Analysis of Teaching Methods in Introductory Physics: a 50k-Student Study," Am. J. Phys. 84, 969 (2016), https://arxiv.org/abs/1603.00516