Originality is something of a legendary beast these days. Or perhaps mythical is a better word. I'm sure there was a time when an artist or author could create something that truly had never been done before, but I don't think that's possible any more. The problem as I see it is simply the volume -- the tremendously vast volume -- of work that's come before. In terms of writing, I think that asshole Shakespeare basically charged ahead and wrote effectively every story that could be written, and he's just one of the thousands and thousands of writers who have already left their mark on the literary history of civilization.
In fact, there's so much out there an author can very easily find himself retreading old ground without even knowing it. I recall a submission I made once to Dungeon magazine. It was a pitch for an adventure arc I was working on. The basic premise was that the party wakes up on a deserted beach amidst the wreckage of a ship, but with no memory of who they are or how they got there. I went on to develop this idea that the party had been evil, and over the course of the adventure arc they would slowly recover their memory and discover that they had been involved in some evil plot, but now they had a chance to start over and work against the forces of evil they had previously been aligned with. Steve Winter, who at the time was editing Dungeon, was kind enough to send a rejection letter (these days you usually won't get a response to a submission at all, so a rejection letter is a welcome surprise). In it, he mentioned that my idea had merit, but it read like the plot of "Lost". Well, I had never seen Lost (and still have not), so this was news to me, but it underscores an interesting point -- that idea I thought was so original was anything but, completely unknown to me. Now that I'm thinking of that adventure idea again, it also sounds a lot like the plot of "Total Recall". So there you have it, originality is tough.
Having said that, I do have an idea on how authors can continue to produce original works. The basic idea is that although everything has been done before, originality can be found in combining ideas in new ways. I like to think of it like a good chef. A good chef isn't inventing new ingredients -- he cooks with the same meats, vegetables, and spices that have been available for centuries. The mark of a good chef, then, is not introducing new ingredients, but rather in combining the old ingredients in new and interesting ways (and, I should point out, demonstrating the technical skill necessary to produce something that's worth eating, aside from whether it's original or not).
So to apply that analogy to RPG adventure design, the thought is that an author simply needs to take some existing concepts and blend them into a combination that hasn't been seen in any other adventure. This has been done many times, of course. Consider the original AD&D module I6: Ravenloft. That adventure stands as one of the greatest of all time, not because it introduced anything new -- vampires were old news even in the early 80s after all -- but because it introduced the theme of Gothic horror to Dungeons & Dragons, something that at the time was a novel concept.
The real challenge is to find some themes and concepts that are not only new to role-playing games, but that can be effectively applied to the concept of adventure design. I don't think I've ever seen a D&D adventure stylized after a romance novel, for instance, and I'm pretty sure I don't want to either.