Tony, [[Wikipedia:Cite sources]] may only be a guideline, but it's inextricably linked to [[Wikipedia:No original research]], which is policy, because the only way you can show that an edit isn't original research is to produce a reputable source.
Doesn't that get back to over-interpreting 'original research'? Which was discussed at length here, a while ago. I distrust statements of this kind, on principle. Anyway it is a poor description of what goes on. A one-line deduction from known facts is obviously not 'original research', whether or not you can cite someone else having already done it. Such things are the small change of doing the research for an article. (Obviously if I read that A is the son of B, I deduce that B is the parent of A; put a few such trivial moves together and you get conclusions which are possibly novel.)
It's easy to make this kind of argument when citing trivial facts. However, most of the original research inserted in these articles is more on the order of "George Bush knew there were no weapons of mass destruction in Iraq, but he lied to the American public and invaded anyway, in order to finish the job his father started, and to restore his family honor".
Jay.