There are a few issues. One is I think you are reading far too much in to the predictive power of SAT/ACT – how well it predicts whether students will fell overwhelmed in college, how well it predicts whether a student will understand the class material, how well it predicts whether the student will feel bored in class, and how well it predicts academic success . SAT/ACT may correlate well with outcome in a small handful of anecdotal personal experiences. However, larger reviews often find much weaker correlations. They also find that the limited predictive ability of SAT/ACT is largely duplicated by other available portions of application.
For example, having a solid foundation is correlated with SAT/ACT, but it is also correlated with taking rigorous classes in high school that are similar to the foundation classes that pre-med students will take during college. If you control for this by comparing 2 students who took the same rigorous high school classes that well prepare for pre-med classes, but had different SAT/ACT scores, then that difference in scores is much less predictive than score in isolation without considering high school preparation for pre-med classes.
This relates to why test optional studies usually find that test submitter kids and non-submitter kids have similar average grades. One of many examples is the Bates 25 years of test optional report at https://www.bates.edu/admission/files/2014/01/25th-Year-SAT-report-Stanford-6.3.11-wch.ppt are below:
Mean SAT Score: Submitters: ~620/~620 , Non-Submitters: ~540/~535
Mean Graduation Rate: Submitters = 89%, Non-Submitters = 89%
Mean College GPA: Submitters = 3.16, Non-Submitters = 3.12
At more selective colleges, where scores have a more limited range with the overwhelming portion of students having high grades, the predictive power of differentiating between students is weaker. For example , the study at https://web.archive.org/web/20100624151156/http://digitalcommons.ilr.cornell.edu/cgi/viewcontent.cgi?article=1014&context=student tries to predict Cornell student grades with a model including HS grades, class rank, ACT score, SAT individual scores, race, hook group, high school characteristics, and many other factors. The full model explained 24% of variance in GPA. The vast majority of variance in college GPA between Cornell students depended on other factors beyond the listed ones. With stats being more compressed today, I’d expect the predive power to be weaker today. Academic success in college and the many other discussed factors depend on far more than stats or how stats compare to peer students. Stats often explain a small minority of variance in academic performance.
.
.
A second important issue is that all colleges do not give the same portion of A grades in pre-med classes. In general, as a large portion of students are well prepared and do high quality work, the portion of students receiving A grades goes up. At more selective private colleges where a larger portion of students are well prepared and do high quality work, the most common grade is typically A. At less selective colleges where a larger portion of students are not well prepared, the most common grade is often not A.
For example, the 2021 Harvard senior survey indicates that the median self-reported GPA of Harvard students approximately 3.9. It’s not just the top 25% of the class who are getting A grades. Is a kid who is not in top 25%, but gets an A still going to feel discouraged because he doesn’t get it in the way you describe? Harvard’s CDS lists a 75th percentile ACT of 36. Would you tell a 34 ACT kid to avoid attending Harvard because his 34 is well below 75th percentile score of 36, so he is unlikely to be academically successful? Should he attend his public safety instead where the portion receiving A grades is far lower?
If a student believes he/she has a weaker high school foundation, rather than comparing test scores to other students, a better idea might be to look at what the college does to support kids with his/her type of high school foundation. Continuing with the Harvard, example, they have all kids take placement tests and a variety to help choose appropriate starting point and level for first year math/science, such that kids who are behind have a good opportunity to catch up. They don’t just throw everyone in the deep end and see who succeeds and who fails.
For example you mentioned calculus being taught at a different level in different colleges. Following the placement test, a Harvard student might start at any of the following math levels – Math Ma,b; 1a,b; 19a,b; 20; 21a,b; 23a,b; 25a,b; and 55a,b. The lowest level (MA) is a half normal speed calc/pre-calc type class, while Harvard’s website describes math 55 as “probably the most difficult undergraduate math class in the country”. They don’t “teach to the level they have coming in,” assuming everyone is at the same level. Many colleges offer similar types of support to help students with relative weaker HS backgrounds, but I wouldn’t assume test score range correlates well with availability of such support.