QA analysts may have reported it. Whether or not there's time for a reported bug to be fixed as of launch is always another question altogether.
At least, that's what I've gathered from what I've read on the subject. Anyone can feel free to correct me. 
It sounds like it's an inconsistent issue, which are always the worst. Especially if it's inconsistent enough to fail "silently." Lets assume that we have 100 people all doing legitimate playthroughs (i.e. not testing with their focus on something else) and something happens 1% of the time. That means that 1 person's playthrough will be affected, and if it so happens that that one person doesn't notice (because they found other issues, or are specifically tasking on something else) then it can get missed, which means we need on average another 100 playthroughs before it occurs again, and we have to hope it gets noticed and not overlooked because some other thing was noticed as well.
Now let's fast forward to release. Lets say a million people all start playing at release. Something that happens 1% of the time is now impacting 10,000 people. That's a lot of people and their concerns are valid. Further, once people starting point it out, then others that might not have noticed also start noticing (this happens in QA too, people say "Hey, keep an eye out for this thing."). I don't think banter has been an issue in my game, but I am definitely much more cognizant of it when I notice "I haven't heard a banter for a while... or am I just going crazy!?"
Numbers just made up for easy math, but it's the crap reality for inconsistent issues
I loathe intermittent problems because it usually means that there's some particularly unusual and not at all obvious set of preconditions (computers aren't really random) to cause it to happen, so now the hunt becomes "figuring out what those problems are." I'd much rather it simply be "banter doesn't happen very much at all for everyone" because then it's easier to identify and fix. Alas.
I do appreciate the feedback though.