Data Guard vs FlashBack for Logical Data corruption

I am looking for a couple of opinions from members of this community who have been doing this kind of stuff probably a lot longer than I have.

Scenario

We have an approximately 1.5 terabyte database (10.2.0.5) with about 1.2 terabytes in readonly tablespaces that we archive off the files once they become readonly then skip on rman backups.  We have a manual data guard setup that we have kept around since 9i setup on the same node effectively being used only for logical corruption protection.. as we don't use it for read only querying and everything is on the same box as we don't have resources for another offsite machine for effective dr. We do use rman to take nightly backups (skipping readonly)