Previously, we had reported on the relationship between AP physics experience and exam scores, and student scores on the Force and Motion Conceptual Evaluation and in an introductory calculus-based mechanics course [E. W. Burkholder and C. E. Wieman, Phys. Rev. Phys. Educ. Res. 15, 020117 (2019)]. Two limitations to our previous work were that we aggregated data from algebra-based mechanics (AP physics 1) and calculus-based mechanics (AP physics C - Mechanics) courses, and that only 25% of the students who took AP physics reported their scores to the university; we addressed these limitations here. We found that there was no reporting bias in these scores for either AP course. We found that AP physics 1 (AP1) exam scores were predictive of students' FMCE scores while AP physics C (APC) exam scores were not--this is consistent with our previous findings. We found that APC exam scores were more predictive of course grades in physics 1, but that students who scored a 5 on the AP1 exam also had high grades in physics 1, despite it being a calculus-based course. These results disagree with our previous findings that AP exam scores were not correlated with final exam scores after controlling for SAT or ACT math scores, likely due to the variability in course final exams from year to year. Though these results are for a single, highly selective university, they provide some key insights to educators about the value of AP courses. One important implication is that students who score a 5 on the AP1 exam can likely receive credit for the calculus-based introductory course, increasing access to higher level physics courses for students whose high schools do not have access to APC courses.