Stuck node detection #108

The controller could track non-progressing nodes and escalate them (e.g. mark the pool as degraded) after a time threshold. This would cover cases where the daemon has not responded to a controller action: desiredImage != booted with Idle=True (daemon should be staging), desiredImage == booted without Idle=True (daemon should have settled), and desiredImageState == Booted with Idle=False reason=Staged (daemon should be rebooting).

Issue originally reported by @jlebon in https://github.com/bootc-dev/bootc-operator/blob/main/docs/ARCHITECTURE.md