add cron check to health check endpoint

We could look at CronWorkflows and determine if there should have been a cron run recently (e.g. in last 5m). If this fails then the workflow-controller is unhealthy and should be restarted.

Start at the health check endpoint and the workflow-controller code, then inspect how CronWorkflows and recent cron runs are represented. Determine how a recent-run window such as five minutes should be evaluated. Done means the endpoint reports the controller unhealthy when an expected recent cron run is missing, causing it to be restartable.

Written by the indexing model from the issue text.