Update monitoring doc with available metrics

[squaremo]

The site/monitoring.md doc is way out of date. These are the metrics I can get from the /metrics endpoint of the daemon (I've truncated the timeseries under each to just one; assume a count where the metric is a histogram):

# HELP flux_cache_request_duration_seconds Duration of cache requests, in seconds.
# TYPE flux_cache_request_duration_seconds histogram
flux_cache_request_duration_seconds_bucket{method="GetKey",success="false",le="0.005"} 2315

# HELP flux_client_fetch_duration_seconds Duration of remote image metadata requests, in seconds
# TYPE flux_client_fetch_duration_seconds histogram
flux_client_fetch_duration_seconds_bucket{kind="metadata",success="true",le="0.005"} 0

# HELP flux_daemon_job_duration_seconds Duration of job execution, in seconds.
# TYPE flux_daemon_job_duration_seconds histogram
flux_daemon_job_duration_seconds_bucket{success="false",le="0.1"} 0

# HELP flux_daemon_queue_duration_seconds Duration of time spent in the job queue before execution, in seconds.
# TYPE flux_daemon_queue_duration_seconds histogram
flux_daemon_queue_duration_seconds_bucket{le="0.1"} 4

# HELP flux_daemon_queue_length_count Count of jobs waiting in the queue to be run.
# TYPE flux_daemon_queue_length_count gauge
flux_daemon_queue_length_count 0

# HELP flux_daemon_sync_duration_seconds Duration of git-to-cluster synchronisation, in seconds.
# TYPE flux_daemon_sync_duration_seconds histogram
flux_daemon_sync_duration_seconds_bucket{success="false",le="0.5"} 1

# HELP flux_registry_fetch_duration_seconds Duration of image metadata requests (from cache), in seconds.
# TYPE flux_registry_fetch_duration_seconds histogram
flux_registry_fetch_duration_seconds_bucket{success="false",le="0.005"} 39