An evolving body of observational data now support at least a biphasic course for the natural history of COVID-19. Evidently, about 80% of infected people recover from self- limiting illness, suggesting temporally appropriate innate and subsequent adaptive immune responses conferring the important ‘antiviral state’. Initial viral evasion of the host innate immune response seemingly precedes a delayed, exuberant inflammatory reaction, which combined precipitate the ‘hyperinflammatory state’ in severe infection. Intriguingly, the late-enhanced immune response has a myeloid signature (likely IFN-I induced) involving inflammatory monocytes, macrophages, and neutrophils. This corroborates pathophysiology of the related SARS-CoV-1 bearing similarity to macrophage activation syndrome and observed lung immunopathology of COVID-19 characterized by ARDS. Collectively, these point to a still undefined inflection point beyond which the host response is inadequate to contain the virus and the hyperinflammatory state ensues. If this model for COVID-19 holds true, a ‘double-hit’ is perhaps the ideal therapeutic strategy. Specifically, the 1st hit must be virus-directed and target replication with a backbone antiviral, e.g. a repurposed or structure-based nucleoside analog; the 2nd hit must be host-directed to counteract initial viral innate immune evasion or subsequent hyperinflammation. In this way, we hypothesize that phase-specific therapy may hold the best promise for COVID-19.