Is it OK to use the serviceescalation Nagios facility to reduce frequency of the notifications sent?
The idea is to setup a generic service template with frequent notifications by default and provide means to automatically reduce Nagios' peskiness when the particular service' outage is ignored long enough.
Any drawbacks to this approach?
It is your process. Use Nagios to best implement it. The flexibility of Nagios shines in use cases like this. Make sure everyone understand the implications and go for it.