Out of the box you have alert rules and performance collection rules and event collection rules. Alert rules fire in response to script based data and or event logs, or log files, etc.
The overrides for disk space, cpu, those come from monitors. Locate the monitor, and look at the properties of the monitor. This will show what management pack the monitor is from, and what TARGET the monitor is using.
If you want to modify monitoring for a select group of disks or cpu's, then you are going to have to create a dynamic group of said objects using something like a regular expression to filter based on server ou, name, etc. When you have a group containing the targeted object class, then you can override the monitor using the instance group and set overrides based on what are expected from objects in those instances.
So when creating overrides, pay attention to the TARGETED CLASS of the monitor.