Fleet-wide host config drift detection for Scaleway Apple Silicon machines

The CAPI Scaleway Apple Silicon provider now detects drift using a fleet-wide hash of all host configuration it pushes, including install scripts, firewall and VM networking setup, launchd jobs, and embedded binaries such as tailscale and node_exporter. When the operator starts up, it computes a single canonical hash from only the operator image and fleet-level inputs, with all per-host fields zeroed, and stores it on each Machine’s status. If it differs from the operator’s current hash, the provider re-pushes the host config automatically on the next reconcile. This means script fixes, fleet CIDR/tag/route changes, and updates to pushed binaries now roll out to already-provisioned Mac hosts without manual SSH intervention or force-patching Machine CRs. Existing machines will re-push once,