Refactor pending_* tasks in order to output more focused information

Currently the only difference between the two tasks is the prometheus query being used for identify the hosts that should be looked up.

Both tasks output the same information which is confusing. When we ask specifically for pending reboots, we don't want to know about which services are in need of a restart.

The output is also changed to remove the duplicated information showing the full list of hosts and the full list of services. We're already seeing that information when listed per machine.