Do we need empirical evidence to warrant thinking that telling children that people suffer torment in hell forever is harmful and bad? I don't think so.
I disagree with this - unless she means: do we need more empirical evidence (more evidence, that is, than we already have on such issues)? But if Ophelia means, as I take her to, 'Do we need evidence for thinking that hell-talk harms children?' then yes we do. It's an empirical hypothesis, and like any other such requires the support of evidence. The fact that Ophelia goes on to say we can in this matter extrapolate from experience (among other things) points in the same direction. The experience we have contains various forms of evidence. Furthermore, in principle we have to allow for the possibility that new evidence might show - though I don't, myself, believe this is likely - that the beneficial effects on children of hell-talk outweigh the harmful ones. Could be, you know, that it toughens kids up and better prepares them to meet the harshness of the world. Unlikely, as I say; yet, although there are claims that don't depend on empirical evidence - such as that it's wrong to cause unnecessary suffering - I can't see that a claim (of fact) about what harms people can do without the support of such evidence.