Retrieving fails with 400 errors from API

I'm getting this continuously in the log:

time="2018-05-22T18:37:40Z" level=error msg="Error retrieving Time Series metrics for descriptor `loadbalancing.googleapis.com/https/backend_request_count`: googleapi: Error 400: The provided filter doesn't refer to any known metric., badRequest" source="monitoring_collector.go:176"
time="2018-05-22T18:37:40Z" level=error msg="Error retrieving Time Series metrics for descriptor `loadbalancing.googleapis.com/https/backend_request_bytes_count`: googleapi: Error 400: The provided filter doesn't refer to any known metric., badRequest" source="monitoring_collector.go:176"

Running with the command line:

--monitoring.metrics-type-prefixes loadbalancing.googleapis.com,storage.googleapis.com
--monitoring.metrics-interval 30m
--monitoring.metrics-offset 2m

I'm not sure if it's just a metric that doesn't exist (e.g. an LB without HTTPS?) and that this is ignorable, but it's logged as error.

Version 0.5.1.

Start at monitoring_collector.go:176 and reproduce the 400 errors with the listed monitoring metrics flags and version 0.5.1. Check how unknown or unavailable metrics are handled by the Stackdriver API, then verify that the resulting behavior and log severity match the intended treatment of missing metrics.