Refactor initial static deployment through MainDeployer/URLDeploymentScanner

An issue related to the started notification discussion in the linked thread, is the behavior of the URLDeploymentScanner emitting a Incomplete Deployment exception on every scan that has services that have not started. The problem here is that if a groups of services is going to be delayed from starting until the server sends its started notification, the URLDeploymentScanner will complain about failed services. The startup and deployment layers need to have a better understanding of the possible multi-stage, system V run level behavior to not emit spuriours error messages.