At the moment it appears that to determine group status, for a group to be down you can either have one member down for the whole group to be classified as down (Show worst status), or you require all members to be down (Show best status and Mixed status shows warning).
It is possible to create a child/parent group scenario to mix and match, however this still doesn't cover all scenarios. For example, within a large server farm the service may only be degraded or down when 51% of the servers are down at once.
Because of dynamic queries determining group members, I don't see it as being possible to assign a group critical flag to an individual device, however I do see a situation where you may have a parent group that defines a service that is based on Show worst status, with children groups beneath.
You would have a Show worst status child group where all the devices are critical to the service, a Show best status child group where the devices are clustered or in a failover situation, and an X% down child group where a certain percentage of nodes down (load balanced or server farm situation) will cause the service to fail.
This would allow for better monitoring and reporting of complex services. If I'm not clear in explaining this or if you require more info please let me know.
I'll mark this for the PM.