One problem I see with this debate over "original research" is that we are trying to apply it as a uniform rule to every possible case, when in application it will take many forms.
Take, for example, David Gerard's example of Indie Australian bands.
From my reading, most of our articles about music bands mostly set
forth verifiable facts such as when the date the band was formed, who was in it, & what records it reelased. For matters like these, I have little to no concern about where these facts came from, because one could always find a way to verify them.
But when the article goes beyond these dry facts to the matter of inspiration, the motivation of its members, etc., then I get concerned. As an example, suppose in the article about fictional band "Mestruating Men" the following appears: "After the release of the first album, differences over the band's direction led to personel changes: according to lead singer Wally, "When Harry said he had been influenced by 'Men without Hats' & said we ought to try to sound more like them, we kicked the wanker out of the band & found Bruce to replace him." Now if the quotation came from a reasonably easy source to verify -- say either the Australian equivalent of Rolling Stone or New Music Review -- I would keep it, whether this was original research or not. However, if this came from a primary source that was had to verify -- a personal interview, or a newsletter or publication that is not reasonably accessible -- then I would raise the question of "original research."
(By "reasonably accessible", I not only mean documents that have been published, but can also be expected to be accessible thru a public library; Interlibrary Loan & online databases are tools that every Wikipedian ought to be familiar with. However, some published materials are difficult to obtain: a valuable source for the Seattle music scene for the late 1970s, 1980s & early 1990s -- "The Rocket" -- stopped publication years ago. Although it was a free monthly newspaper you could find at record stores in the Northwestern US, I confess I wouldn't know where someone could find a specific issue of it today -- or if anyone even thought to save any copies.)
I feel this kind of objection ought to apply both to well-known bands like "The Rolling Stones" as well as Australian Indie bands.
Moving to another topic where "original research" will likely cause a problem, let us consider the critical appreciation or investigation as to motives of famous artists or politicians. Obviously, everyone will have an opinion in this area, & doubtlessly we could argue for hours over whether Shakespeare is a greater playwright than Sophocles or Ibsen, & over the motivations of such public figures like Ivan the Terrible, Ieyasu or George W. Bush. Here, there is no good reason for original research in any form: if one knows the secondary literature to a satisfactory degree, then one can simply quote or paraphrase what has been written on the subject. And the average user of Wikipedia will want to know what the published judgement of these people is more than what some Wikipedian thinks.
(And in cases like these, we should cite more than peer-reviewed journals; evaluation of public figures is shaped by more than degreed experts. Poets & actors can potentially tell us more about Shakespeare than any random tenured professor published by a university press; TV Guide's account of Bill Clinton playing the Saxophone on the Orsinio Hall show could tell us more than a Ph.D. working in a political think tank.)
Then there is a third case where we confront original research: in the search for originality of expression, contributors will inevitably introduce their own conclusions drawn from verifiable facts. One example would be from my own contributions, [[Battle of the Catalaunian Plains]]: when I wrote this about 2 years ago, I was very concerned at the time over introducing copywritten text, so I based it on published primary sources, such as _The Gothic History of Jordanes_. Undoubtedly in the writing of this article, I introduced my own POV & my own conclusions concerning the material, no matter how hard I tried not to; we all have conscious or unconscious biasses in the areas we are profess expertise in. However in this case, I submitted it fully aware of the Wiki philosophy, that what gets written may be subject to ruthless revisions. And my own attitude to the article is that if someone can improve on it by citing the necessary experts to confirm or replace what I wrote there, then it is a good thing. Here, calling something "original research" & insisting that the material should be deleted does nothing more than cause friction on Wikipedia, whereas making changes in a reasonable & responsible manner solves the problem -- which I assume is what we all want to do in the first place.
I don't have a clear-cut answer for the issue of "original research" -- except to say that I know it's a problem when I see it. But my point here was to argue that to address this issue we must also acknowledge where it poses no problem.
Geoff