Support multiple scale sets per listener

What would you like added?

I would like to have multiple scale sets but without the extra administrative and resource needs of having multiple listeners (one per scale set).

Why is this needed?

I need different type of runners.
They differ only in their template configuration, everything else is the same (of course their runs-on name should also be different).
The rationale here is to have multiple configurations (different resources, CPU, RAM, storage, different pod annotations (node groups) etc), so we can have different runners for different needs.

Additional context

What we can do now is:

1. create a controller
2. create a listener-GHA scale set pair for all runner configurations

While this works, each listener takes up resources (CPU/RAM and Kubernetes objects, like a ServiceAccount, Secret, Role* etc) and I think conceptually a listener could be listening for multiple scale sets and start the pod with the right template if requested.

Start with the linked charts/gha-runner-scale-set/values.yaml template configuration and trace how the listener and scale set are currently associated. Map how distinct runner templates and runs-on names could be selected, then verify that one listener can support multiple configurations without creating the extra resources described in the issue.