[Fleet] Revise generated policies when using the shipper

[faec]

An impending change in elastic-agent affects how the shipper is enabled on an output. Previously this was done by setting shipper.enabled: true, or similar options described in #141508 (comment). After this PR is merged, the shipper will be enabled on an output by adding the use_shipper: true flag. (Internally this is a more significant change: the previous syntax enabled one specific shipper, named shipper, which was statically configured in the agent repository. The new syntax no longer cares about the specific name of the shipper, and just signals agent to allow any shipper that supports the given output type.)

This will also affect queue configuration. The queue options described in #141508 (comment) were based on plans at the time, but were never implemented -- the current plan calls for queue settings to be included directly in the output configuration block (see the implementation issue). For example the shipper could be enabled with the disk queue as follows:

output:
  default:
    type: elasticsearch
    use_shipper: true
    queue.disk:
      max_size: 10GB
      retry_interval: 500ms

(Ownership of this issue is still being discussed, Agent team might end up taking on the implementation work for this issue.)

[elasticmachine]

Pinging @elastic/fleet (Team:Fleet)

[jlind23]

Shipper project has been revisited since then hence closing.