Probably a combination of them both Corrosion-Fatigue
Most of that link is about aluminium structures. I broke it open to look at the fracture faces and pretty certain it was regular high cycle fatigue. The crack also ran in a slight helix, maybe indicative of the mixed torsional / bending loading in that area (pure torsion would be a 45 degree helix).
That frame had a long hard life – first a training frame for a triathlete (early 90s Hawaii Ironman), then my boss for UK triathlons, then years of singlespeed abuse from me.
All the joints were internally brazed with incredibly small fillets when I cut it up for a look – I was amazed how durable they were – good mitre and proper flow of brazing alloy appear to be the secret. Photo shows the top tube / head tube joint (and a bit of down tube). And once blasted clean there looks to have been very little meaningful corrosion so contrary to hora’s comment steel bikes really are fine for decades of year round use (shock horror).