This is a dreadful article, even if it is an interesting topic.

Their forecasts, and those of many other modelers, vastly overpredicted the actual toll of the disease.

No they didn't.

That prediction was in the absence of interventions.

They modeled the impact of various interventions and found a best would produce around ~10,000 infections[1] in Liberia in November 2014. In reality there were around 6500, so I'd say that was a pretty accurate prediction.

[1] guessing from a graph: http://currents.plos.org/outbreaks/files/2014/09/LiberiaComb.... Complete paper: http://currents.plos.org/outbreaks/article/obk-14-0043-model...