The FDA rarely, if ever, issues an outright rejection on ANDAs.
That's one of those obvious points if you think about it.
For small molecule drugs a broken ANDA would always be fixable because all they have to do is correctly manufacture the chemical and all.
But for more complicated drugs, such as this, I would argue the statement is simply factually wrong. The 3 rejections in this case are enough to be significant over the space.
Back on the main point, I fail to understand why people ignore the possibility that Teva resubmitted with an improved product in face of either a rejection or other bad feedback.
It is the only answer I see that avoids making either DD or RR come out as being very wrong.