The compute operator currently create managedclusterset based on workspace name. We should not only rely on workspace name in case it is deleted and recreated. KCP team is planning to implement persistent identifiers that way we can identify workspace uniquely. Use persistent identifiers for managedclusterset creation when they come.