Hi,
That´s quite a number.
At least what I described here helped us with our CISCO stacks.
I know also from past experience that some CISCO Switches respond very slow after fetching consecutive data. That´s related to IOS Version and model.
There are some blogs from Alexander WILMS describing best practice:
Just as an idea, maybe with help of the rule “SNMPv3 contexts to use in requests” you can distribute the different services over different host. Will not help if almost all services using the same check.
Beside that there is a rule “Fetch intervals for SNMP sections” to cover situations where fetching takes too long.