Accordng documentation Apicast instance should provide these metrics https://github.com/3scale/apicast/blob/master/doc/prometheus-metrics.md
batching_policy_auths_cache_hits
batching_policy_auths_cache_misses
nginx_error_log
nginx_http_connections
nginx_metric_errors_total
openresty_shdict_capacity
openresty_shdict_free_space
threescale_backend_calls
total_response_time_seconds
upstream_response_time_seconds
upstream_status

If we don't consider metrics for Batching policy these are metrics provided by Staging and Production Apicast

List of metrics provided by Staging Apicast:
nginx_http_connections
nginx_metric_errors_total
openresty_shdict_capacity
openresty_shdict_free_space

So missing metrics:
nginx_error_log
threescale_backend_calls
total_response_time_seconds
upstream_response_time_seconds
upstream_status

List of metrics provided by Production Apicast:
nginx_http_connections
nginx_metric_errors_total
openresty_shdict_capacity
openresty_shdict_free_space
threescale_backend_calls
total_response_time_seconds
upstream_response_time_seconds
upstream_status

So missing metric:
nginx_error_log