Skeptics and storytellers
One of the interesting aspects of Quicksilver is that Stephenson has fairly seamlessly integrated fake facts and people into a true historical context. He's incorporating more than historical embellishment; Stephenson introduces the entire Waterhouse and Shaftoe families and inserts them into history -- Daniel Waterhouse as a personal friend of Isaac Newton and Gottfried Leibniz, for example, and Lawrence Waterhouse as a friend of Alan Turing. He also creates the island nation/province of Qwghlm, which I spent two or three days believing in until I thought "... no! that has to be made up."
This reminds me of the old James Michener novels like Space where the author would invent characters and even states (I remember a state named Fremont or something?) who would go interact with important points in politics and national history.
But somehow Stephenson's inventions are so well fleshed-out that they serve as an interesting object lesson in history and research: just because it's a good story doesn't make it true. I have a bad habit of learning history and science from books like this one, but Quicksilver (not to mention Cryptonomicon and -- more to the point -- Snow Crash) serves as a lesson in not trying to pick up facts just by osmosis. (Snow Crash in particular has some very bad science in the linguistics digressions, but it's very appealing -- I have a story about a late night with a bartender who wouldn't be convinced it was malarkey that I'll tell later.) But I find myself really wanting to fact-check his ass. He's a smarty-pants who feels no compunctions about dropping in fake people, nations, or facts about the universe just because it's a good story. I'm often guilty of the same thing, but he does it so brazenly that it's kinda fun to spot them and say "how could he think he could slip that one by?"
I imagine a reading group or even a high-school lit or history class that might tear its way through one of these books trying to figure out what parts of it are historically justified, what parts are possible but for which there's no evidence, and what parts are demonstrably false. Woo -- wouldn't that be a fun class? You could read Quicksilver and call it "History of the Scientific Method/Science applied to History", and take the story itself as the subject of the method.
Actually it would be more appropriate if you cut out all the sex bits from the novel, which -- quite frankly -- aren't very sexy and seem to be in there by contract with the publisher ("Mr. Stephenson agrees to put in at least one discussion of sex per chapter, and at least every other one will involve painful innuendo and/or implausible attraction or description of sexual organs. The publisher agrees not to edit these sections.") Cutting these out would shorten the novel by ten to fifteen percent, making it lighter-weight and easier to inflict on unsuspecting highschool students, because it would avoid shocking parents, it would improve the novel, and it would stick to what Stephenson does well -- making 17th century science politics readable and interesting.
Oh, and chase sequences. He does those well too.