Logs/diagnostics for HTTP Health checks
There seems currently no way to know whether an HTTP health check is correctly configured, whether the health check requests are being issued, or if they are and the responses are failing, or what the error responses are.
For example, in my case, I forgot to add the firewall rule that is needed to allow health check requests to reach the VM. And trying to debug the app, or http health check config is futile. You have to recognize requests aren't coming through, then read the docs carefully to know about the firewall rules.
Instead I'd expect a more useful (than the relatively useless info there currently is about a health check being created/deleted) in the log for an http health check to record what the health checking system knows - the specific failures causing failed health checks.