I'd say, if the behaviors and thoughts are interfering with life activities or causing suffering over some months or years, it's time to investigate.
I think this is key. I don't actually see what it would mean to answer the question in the subject line - we could draw a line and say that side is a sensory OE that side is SPD (for example) but I think such a line would always be arbitrary - we should draw in in a place that is pragmatically useful in getting the right things done. You don't want either to overreact to an unusualness (label something as 2e that's better treated as OE) or underreact (vice versa).
I'm sorry you lost years of treatment; sounds like the line didn't get drawn in the right place in that case. Here's a different anecdote:
When DS was 6 and struggling with what I labelled emotional OE, I found it useful to draw his teacher's attention to the notion, because this helped her to see his behaviour as a curiosity she could work with and around rather than as a disorder she should hand over to some other professional. Part of what I was concerned about, actually, was the risk that DS might get diagnosed with Asperger's, something I felt wouldn't be correct or helpful. (And in fact, reframing it for that teacher helped tremendously, and now that DS is 8, I don't think anyone would remotely consider that he might have an ASD; in that sense, I'm confident I made the right call.)
The behaviours and feelings that are labelled OEs certainly exist in some children; I suppose an interesting and answerable question would be, are they actually correlated with giftedness? You would need a careful study design to be sure, with good definitions of both what it is to have an OE and what it is to be gifted, avoiding being either more or less likely to count a child as gifted as a side effect of them having OEs, for example! But it should be doable. I'm not aware that such a thing has been carefully done? My hunch is that you might get different answers for different OEs, but that you'd find something interesting. Of course, it could still be a useful notion even if there turned out to be no correlation with giftedness.
