Or that DAA mechanics are too unknown - as Dew pointed out, the 'strange' timing of relapse after EOT could really throw a monkey wrench into the timing of the trials or the approval cycle. E.g. if SVR12 is a reliable indicator of efficacy after 12 week treatment then you can tune your treatment after a trial of only about 1 year duration. I.e. in 3 years you get maybe 3 cycles of different DAA groupings, ... . But if you are forced to a 52 week SVR (or worse? - to prove that x weeks really is adequate) you get a little less than 2 cycles in 3 years. Plus there is just the delay due to regulatory uncertainty.
Keep in mind this late relapse, in the ABT study, was seen in a oral combination which didn't have a nucleotide analog backbone. No such event has been documented using GS-7977.