Splits from IR-490 
As a result of the review of what is/was offered with https://github.com/openshift/cluster-image-registry-operator/pull/1147 it was noticed by Product Operations that the metrics being presented here are to highlight when something catastrophic happens (and things are broken). However there isn't an 'early warning indicator' for either alert (primarily storage is filling up) 

Thus it is suggested that an IR metric to check if an Image Registry (IR) is near a defined  storage limit threshold (default to say 80% if a user doesn't define an alert threshold), to warn users about impending issues; would be a good addition to the system.