"I think this obsession with “perfect” standardized test scores is silly. It’s easy to look at the score distributions and say, “Wow, only 2,760 kids scored a perfect 36 on the ACT in 2017, but a whopping 12,836 scored 35—so the 36s are in the top 0.1% (actually the top 0.136%) while the 35s were merely in the top 1% (0.697%). That’s an order of magnitude difference, so the perfect 36s must be truly the academic elite, while the 35s are merely excellent, but on a distinctly lower plane.”
You’ve described perfectly my problem with US standardized tests. They simply don’t have enough discrimination to capture the long tail instead of the random fluctuations of performance within the top 1%.
For comparison, the typical math test my UK company used for potential employees had 32 questions to answer in half an hour (these were not intrinsically much harder than a math SAT-2 but more complex questions in way less time). If you scored 19 out of 32 that was benchmarked to be the top 5% of U.K. college graduates. Top 1% was 23-24 out of 32. The best score I know of (in a company with many Oxbridge graduates) was 29 out of 32.