Personally, I have nothing against new approaches - but to use something new just because it is new is not a justification for doing so. Somebody's proposition that a new method has merit is NOT proof that it does.
Why do you assume they don't have a good reason to switch teaching methods? They're Scientists, after all. Also, someone has to test new methods that look promising theoretically. Some methods are inevitably getting scrapped. That's normal and good.
That shouldn't make anyone assume the scientists don't know what they are doing. (Not that I think you are)
Educational science is to science as sewer maintenance is to hydrodynamics.
How do we know? Look around! Achievement levels upon HS graduation are at all time lows (specifically in the US). Sure, there are many who succeed in spite of the system but the vast majority graduate dumber than a box of rocks.
I prefer to think that the teachers want an elitist system whereby they seem to be the smartest person in the room. Big words, complex ideas (badly explained), parents unable to help with homework... Yup! The teacher likely is the smartest person in the room. At explaining some convoluted approach to mathematics.
Teachers, typically, make squat money. Society doesn't value them, just ask. We don't value them (and I'm talking at teachers less than full-blown professors at prestigious universities) because their performance (teaching kids to enumerate) sucks.
Did I mention that they don't want to have their salary tied to performance? I realize how complex performance measurement can be when they inherit a class full of kids who don't want to learn. I don't know how to deal with that. We have kids whose attitude is "Nobody in my family ever got an education and I'm not going to get one either!". But they are required by law to plug up the system. That I don't now how to deal with it doesn't mean there isn't a solution. In terms of percent improvement, it's easier to show this when the average is quite low. What you don't want to be measured on is a bunch of AP students already at some 90+% level. Percent improvement is hard to get!
You might notice that I don't hold our educational system in high regard.
New methods? Why not use the same methods that created the greatest achievement in the history of mankind, landing a man on the moon and returning (landed July 20, 1969, almost 50 years ago). We did that with conventional methods and extraordinary people but we did it with slide rules and times tables. Does everybody remember that we made 6 landings and 12 astronauts have walked on the moon's surface? Keep that in mind when comparing methodology.
It's a lot like the Calculus course my grandson is taking. A very small percentage of the material will ever be applied outside of a math department. Limits aren't a big deal, finding concavity is just about useless and proofs of L'Hopital's rule just never came up in my corner of engineering. What did come up was related rates and optimization - the word problems. Sure, the subjects are taught but only as a side issue to the 'elegance' of math. Who cares? If it doesn't solve a real-world engineering problem, skip it and move on! Have a second track for math majors. I understand (from a video course) that the University of Florida does indeed have two tracks through math - one for math majors and another for engineering majors. Good for them!
At the end of my rant, I guess I just don't consider those people tinkering with teaching math as scientists. More important, I don't want my grandson to become a lab rat.