At the GURPS session of today's Fnordcon, there was a question about how much research a GURPS book required. The answer is, of course, that it varies with the length and nature of the book, but I'd like to point out another key factor: the author.
I write historicals in large part because I've got a social science and humanities background. I'm already oriented in various historical periods and in the language of history, archaeology, and anthropology. There's basically an investment of research humming along under the surface there. Consequently, I do less research specifically for any given work than one might think. Indeed, there have been some things where there's a specific place or idea that comes up in my reading which looks like a good idea for a GURPS work and only a fraction of the resulting work comes from research specifically for it. The rest is contextualizing and providing background from stuff I already know.
Consequently, I can write something in greater depth than somebody without that background in the same time. If, say, Peter or Christopher were to have written GURPS Hot Spots: The Silk Road instead of me, he'd probably have ended up doing a lot more research than I did. Similarly, if I wrote GURPS Martial Arts instead of Peter...yeah, I probably wouldn't be done yet.