Good question. Doing a /host=*/subsystem=logging is certainly one way to deal with host-controller.log without adding new code. Perhaps overkill though. It would need to be a Feature of its own though, with this log viewer aspect as just a detail.
However, process-controller.log would not be technically related to any /host=*/subsystem=logging. IOW, what's in such a subsystem would have no impact on the logging behavior of the ProcessController, which is a separate process not managed by the HC. Perhaps the host subsystem could still serve the file though so long as it's in the same dir where the HC writes its logs. I forget what the precise rules are on what files the subsystem log viewer is willing to serve.
At some point we should probably consider adding the logging subsystem to the host resource. I thought there was already a JIRA for that, but I can't seem to find it. However it might be best to wait until we have a solution for removing the logging.properties file.
That said the idea for the host-controller.log was to reuse what is in the logging subsystem for reading log files. I'm not too sure what could or should be done with the process-controller.log. I haven't put much though into. Currently the HC and PC logs can only be configured via the logging.properties file. This may always be the case with the PC.