Wrong cleanup of stale conditions from OCPBUGS-2783

While updating a cluster to 4.12.11, which contains the bug fix for [OCPBUGS-7999|https://issues.redhat.com/browse/OCPBUGS-7999] (which is the 4.12.z backport of [OCPBUGS-2783|https://issues.redhat.com/browse/OCPBUGS-2783], it seems that the older {{{Custom|Default}RouteSync{Degraded|Progressing}}} conditions are not cleaned up as they should, as per [OCPBUGS-2783|https://issues.redhat.com/browse/OCPBUGS-2783] resolution, while the newer ones are added.

Due to this, on an upgrade to 4.12.11 (or higher, until this bug is fixed), it is possible to hit a problem very similar to the one that lead to [OCPBUGS-2783|https://issues.redhat.com/browse/OCPBUGS-2783] in the first place, but while upgrading to 4.12.11.

So, we need to do a proper cleanup of the older conditions.

4.12.11 and higher

Always in what regards the wrong conditions. It only leads to issues if one of the wrong conditions was in unhealthy state.

1. Upgrade
2.
3.

Both new (and correct) conditions plus older (and wrong) conditions.

Both new (and correct) conditions only.

Problem seems to be that the stale conditions controller is created[1] with a list that says {{CustomRouteSync}} and {{DefaultRouteSync}}, while that list should be {{CustomRouteSyncDegraded}}, {{CustomRouteSyncProgressing}}, {{DefaultRouteSyncDegraded}} and {{DefaultRouteSyncProgressing}}. I read the source code of the controller a bit and it seems that it does not admit prefixes but performs a literal comparison.

[1] - https://github.com/openshift/console-operator/blob/0b54727/pkg/console/starter/starter.go#L403-L404