The current data model for aspects assumes that all subjects have the same thresholds for that aspect. In the example from the docs, unemployment may be OK between 0-7% for USA but for Egypt, maybe 0-10% is OK. This isn't possible to implement using the current data model without duplicating the Aspect for each subject (e.g. Unemployment_USA, Unemployment_Egypt, Unemployment_Mexico, etc).
One way that may be possible while maintaining backwards compatibility would be to allow a Subject to override the critical/warning/OK ranges for particular named Aspects.