I know your question is trying to squeeze a fine point of regulation - but often the situation is actually simpler in practice because if there's that much porridge you're probably off to your #1 div anyhow.
Admittedly the situation is slightly contrived but not impossible. In fact I corrupted the following real METAR:
METAR ESGP 052120Z 00000KT 9999 R01/P2000 R19/0400V0650 BCFG NSC 15/14 Q1018=
Goeteborg does have an NP IAP (NDB/DME) to 01 and I thought with NSC the obvious answer was a visual approach to 01, so applied a ceiling. But the scenario is not all that far fetched.