Add a metric for if all the config checks are healthy

We should add a metric that exposes the config check health as a metric, so CM could alert admins if Hue started unhealthily, or eventually became unhealthy due to an upstream service being down.