Try telling that to the students who have had an appalling low standard of education because of the 90-99% failure rate of all the new things they had tried on them.
I have not read any studies which claim a significant number of these new techniques are creating an appallingly lower standard of education than the students would have gotten otherwise. In contrast the most damning criticism is usually that they techniques had no effect. In my opinion, this isn't because the new techniques are that good. It is because it is really hard to do worse than the status quo.
Nobody would accept a 90-99% failure rate for medical innovations which get as far as being tried on patients!
Depends on the possible side effects and depends on the most likely outcome using conventional medicine. If I have a 100% of dying, a 2% chance of success is starting to look pretty good. And if the worse thing that could happen is a little diarrhea, a 10% chance of completely curing a disease also sounds really good.
there is really no way to determine whether a new technique is effective other than to try it on students. So while education is more risk tolerant than medicine it is nowhere near as risk tolerant as VC industry funding.
Many new educational techniques can be tried out in a very agile manner. Even conventional education tries new ideas constantly, just usually with a less scale-able and less ambitious approach. Initial trials of a single lecture or single lesson plan, measured with a single test, can provide initial indications of success at scale.
Just as you don't test a new CRM app with a large scale deployment at a Fortune 100 company, you don't have to test a new educational technique by changing an entire State's teaching standards over night. All of this can be, and already is, tested on a much smaller scale.