Scott Hermes Thank you for posting your question in Microsoft Q&A. Based on my understanding you observed high-Capacity usage in Primary vs Secondary which has higher number of requests.
As described in docs: What is capacity, in general it might be higher due to number of factors such as connection patterns, size of requests/responses, complexity of operations etc. Or it could be Platform updates such as OS updates on the compute units as well which would contribute to higher capacity usage. It is also important to note that Capacity is not a direct measure of number of requests being processed and it could even spike even if no requests being processed due to the updates.
Since you mentioned it is same API, I assume it has same policies, complexities etc. for similar requests and you see high capacity for more than 7 days, I would recommend you open a ticket with our support team to investigate the backend logs. They have the best tools, resources to find out the cause of the issue.
If you don't have support plan, let me know and I will assist you in creating one-time free technical support ticket. I hope this helps and let me know if any questions.