It is certainly not an insurmountable technical challenge to develop age-based norms for achievement tests. (In fact, nearly all of the reputable individually-administered achievement tests have such norms.) However, there is a perspective that finds that comparing students to the peer group that has received the same instruction is more equitable, and a more accurate reflection of level and rate of academic learning. I agree that doing so introduces possible inequities in the selection process, but using age norms can too, if one assumes that some younger students will perform less successfully simply because of more limited exposure to content and direct instruction in skills. Or that some students who are old for grade will perform less well on age-norms because they have received less extensive instruction than their age- and ability-peers who happen to have been placed into a higher grade level (e.g., in states where age-of-entry varies from district to district, a transfer student from a district with a later cutoff into a district with an earlier cutoff may end up a grade ahead of a same-age peer).
There really is no perfect solution for students on the cusp.
