In most districts, MAP is used for universal screening and progress monitoring of low and at risk performers. Universal screening obviously means every student must be tested, even those clearly not at risk. Any benefit that accrues to high performers often is incidental. Its use for access to higher level programs is data-driven only at a fairly low cut score of 90th %ile.
Remember that most of the children we discuss on this forum are singularities in their schools. MAP is quite useful for the vast majority of students. From the school's standpoint, there are equity issues any time one starts exempting any category of student from an universal screening measure. Also, pulling out the extremely high performers may skew the norms over the long term, since NWEA uses actual test takers to develop future norms.
To your actual point regarding using available data to inform instruction: yes, it would be optimal if these data affected placement. One should also note, though, that schools do experience gaps between how students test, and how they perform in classes. These gaps may occur for various reasons, but one of them is that group standardized testing doesn't measure all the skills needed to be successful in a course of the indicated grade level. (Many of these are not academic skills, but soft skills.) Schools are often afraid of misplacing a student based solely on test scores, resulting in negative outcomes for both student and school, as nearly every school has experienced this at some point. Reading and language in particular can be challenging, as it is quite difficult to get a good measure of writing skill from a computerized multiple choice exam, yet writing is a huge part of language arts classes.
