Oh I don’t have experience teaching a second language but I have quite a bit of experience teaching programming. Luckily, when I’ve taught programming it’s been in a context where there is no need to validate the learners acquisition outside of their own satisfaction.

In a school setting however, there is a strong urge by both the students and the teachers and the external system to provide some sort of “proof” that a standardized level of comprehension has been achieved. Students want to get As, teachers want reproducible methods, and administrators want standardization.

My intuition is that language acquisition, more than almost any other subject, has to be done naturally, idiosyncratically, and non-linearly. This is horrible from the perspective of needing to provide a test at the end of the class.