We've had this discussion before, but I'd just like to point out the other two (well, 1.5) sides of that: there are some (how many would be an enlightening stat) flying saucer proponents who use the term UFO to mean (or at least imply) flying saucers. Some may be careless, some may be ignorant, and some certainly are dishonest. For a 'nobeliever', a judgement has to be made which they are dealing with.
And lets not forget: using the term "correctly" or not, the reason the discussion exists at all is because people are at least hoping that when the FO gets Id'd, it is positively shown to be a flying saucer. As I've pointed out before, if people were hoping to find new atmospheric phenomena (for example), they'd just call themselves meteorologists.
In any case, if you really want to be strict about the definition, you should probably consider the history and entemology. The term was coined during Project Blue Book as a replacement for 'flying saucers' (regardless of who actually coined it, it was coined for Project Blue Book, which had the specific purpose of evaluating their threat to national security). It is a better word, to be sure, but the implication that the term is referring to possible/suspected alien spacecraft dates to its origin and because of that, even those who seem like "respectable" ufologists often use the terms interchangeably. http://www.ufoevidence.org/documents/doc365.htm
Interchangeable usage of the term:
Characterization of the purpose of ufology via the questions it is to answer:
It is fitting though - an ambiguous term for an ambiguous pursuit.
The guy who wrote the article also understands me pretty well:
Well - by definition or by statistics. Assuming the stats weren't falsified for the purpose of a coverup, Project Blue Book judged 94% to be identifiable as non-flying saucers (I'm not sure about the level of certainty required for such a judgement) and judged 0% to be alien spacecraft. So since there were roughly 12,600 reports, a starting assumption that a new report is probably explainable has a 94% probability of being right (caveat: on this site, such reports are typically pre-filtered) and "not identifiable as a flying saucer" assumption has a 99.992% probability of being right. And that's the reason why such starting assumtions are reasonable - just as reasonable as a starting assumption that a new gravity experiment will produce results that conform to GR.