Option B makes sense here since resource groups are really just for organizing resources, not making them resilient to datacenter failure. For real availability if one Azure datacenter goes offline, you'd want to use Availability Zones. I think B is the right call, but if anyone's seen situations where resource groups help with HA, let me know. Agree?
Resource groups are for organizing resources, not for providing high availability across datacenters. To get protection from a datacenter failure, you need to use Availability Zones, not just split things into different resource groups. Clear scenario, makes it easy to spot why B is right.