Regarding the portion of class with a 4.0 GPA, the most recent internal publications are for the class of 2016 at https://admission.stanford.edu/apply/selection/profile16.html . It lists the following
58% of applicants have 4.0+ GPA and 75% of admits have 4.0+ GPA
28% of applicants have 3.7-3.99 GPA and 21% of admits have 3.7 to 3.99 GPA
14% of applicants have below 3.7 GPA and 4% of admits have below GPA
Comparing to the CDS for the same year at https://ucomm.stanford.edu/wp-content/uploads/sites/15/2018/06/stanford-cds-2016.pdf , it shows similar numbers
Website – 96% of admits have 3.7+, 4% of admits have below 3.7
CDS – 93% of class has 3.75+, 5% of class has below 3.5-3.74, mean = 3.94 GPA
Looking at later years of the CDS
2016 CDS – 93% of class has 3.75+, 5% has below 3.5-3.74, mean = 3.94 GPA
2017 CDS – 95% of class has 3.75+, 4% has below 3.5-3.74, mean = 3.95 GPA
2018 CDS – 95.6% of class has 3.75+, 4% has below 3.5-3.74, mean = 3.95 GPA
2019 CDS – 95.4% of class has 4.0, 4% has 3.75-3.99, mean = 3.96 GPA
Note that between 2018 CDS and 2019 CDS, the GPA categories changed, but the percentages did not. In previous years ~95% of the class had a 3.75+ and ~4% had a 3.5 to 3.74. However , in 2019 ~95% had a 4.0 and ~4% had a 3.75 to 3.99 . So I expect that there is a typo in the 2019 CDS actual the actual percentages were as follows.
3.75+ GPA – 95.4% of students
3.5 to 3.74 – 3.6% of students
3.25 to 3.49 – 0.9% of students
3.0 to 3.24 – 0.1% of students
One should also consider than 18% of students were not included in the GPA calculation, and I believe this is recalculated GPA in which freshman year and non-core classes are not included, and A- is treated as a 4.0.
You can see some example students in the decision threads on this site. I expect CC decision thread posters have higher stats than typical, yet a good portion of unhooked admitted CC posters do not have a 4.0 UW. They almost always have a very high GPA, but not 4.0 UW, prior to recalculation.
Regarding the original poster’s question, an A in calc AB (sophomore year) followed by an A in calc BC (junior year) is almost certainly going to be viewed more positively than getting a B in calc BC followed by no math in junior/senior year.
However, there are a lot of other factors. For example, how do you know he is going to get a relatively low grade in calc BC as a sophomore, but not in calc AB as sophomore or calc BC as junior? This sounds like a lot of guessing. If he is taking calc BC as a sophomore, I expect he is an exceptional math student and is not struggling in math. What has he done to suggest that he will not get an A in math?
I was accepted unhooked to Stanford several years ago when Stanford reported ~70% of students had a 4.0+ GPA, rather than the 75% today. Like the OP, I took calc BC as a sophomore (as I recall), and went on to take higher level math and other colleges at a nearby university. I did well in math and the university classes.
I did not do as well in my HS English classes and some of the HS language classes, so my HS GPA was well below 4.0 . However, my GPA showed a good upward trend . As I recall, the most recent grades on my application included:
Math, Science, and Electives – A’s in university classes at likely higher math/science level than anyone from HS had ever taken before, with great LOR from math professor
HS English – A (did much better after switching from AP English track to guided independent study English)
HS Language – A+ (did much better after switching from Spanish to Latin)
I expect that Stanford considered my grades in context. I was a prospective engineering major, so there was a greater focus on my math/science strength than my relative weakness in English/language – both in the grades achieved and in rigor/level of courses. The upward trend was also considered, with a greater emphasis on the more recent courses. As mentioned earlier in the post, Stanford recalculates without freshman year.