PIs were also must have for HIV until they weren't I agree that NS5A is certainly one way to go, and they are ubiquitous, potent, and potentially high resistance barrier w next generation, but certainly not mandatory for a cocktail (sofo and olysio for example was quite good)
I'm of the opinion resources are better spent on NASH if they really have something worthwhile cooking in the lab - there's just so little room for improvement in HCV and you have to get there reasonably quickly (and I could only see breakthrough in HCV now for current DAA failure pts - for whom nuke or CI w PI +/- NS5A best)