Expanding on what I wrote in a comment elsewhere, I think a crucial point to emphasise is that an ideal answer will make both prescriptivists and descriptivists happy.
A good prescriptivist answer will appeal to the most authoritative authority available; and for the most part, the big dictionaries and the more modern style guides are actually pretty descriptively well-informed these days. An un- or badly-sourced prescriptivist answer can always be battled on its own territory by “I trump your Strunk and White with my Fowler!” (or CMoS, or OED, or…); and then if the prescriptivist rises to the bait and questions the claimed higher authority, they can hardly avoid getting into serious linguistic questions and hence on to more descriptivist ground (cf. @nohat’s answer).
On the other hand, an ideal descriptivist answer will not just say “this is what people use these days, and I’m a young native speaker so I should know”; it’ll say a little about what’s used where, and back this up, so will usually look something like
In formal writing [or “in UK English”, “in traditional usage”], X is standard (cf. Fowler); but Y is more common in casual usage [or “in the Southern US”, “in younger speakers’ usage”, …], as can be seen in e.g. COCA.
And as long as these distinctions are clear, even an arch-prescriptivist will be quite happy with this: he’ll see the “correct” usage in the most prestigious place, backed up by a respected authority, and can even enjoy some schadenfreude at the documentation of how badly the rabble are speaking these days!
However, all these are just prescriptions(!) to follow in writing your own answers. When it comes to dealing with other people’s bad answers, I’m not sure what to add to what @nohat said — and yes, it does sometimes get tiresome!
