I'll echo the diagnosis—creating an adventure path is a risky, intimidating, and expensive proposition. It also pretty much requires a rapid release pattern; Shackled City came out irregularly in its first incarnation, and that was a detriment to it since the It's also got a pretty steep learning curve—I've learned a LOT about how to set them up and to orchestrate their creation over the past 7 that I've helped launch (and that makes me feel old realizing I've developed/edited/helped write well over half a dozen of these things!), and having just about finished the development of the start of the 8th and done the brainstorming/plotting for Paizo's 9th one last night, I'm STILL figuring them out.
As for how many can the market support? I'm not sure; we're not there yet. Pathfinder's still going strong, and we're certainly outpacing most folks' ability to keep up with actually playing the adventure paths... but here's the secret. They don't HAVE to be played to be successful. If a gamer buys an adventure path and only reads it, that's a success as far as I'm concerned. Even if he never runs it. Even if he never actually uses elements from the volume in a different game he runs. There's a strange disconnect, I think, from gamers who assume that if they never actually use content they buy directly in a game they're running that they're not getting their money's worth. I don't agree with that at all; if you spend 30 bucks on the new Stephen King hardcover and only read it, you get your money's worth there. For a lot of people, it's just fun to read game books. I know I'm one of those people; I have shelves and shelves of hundreds, probably thousands of game books I've collected over the years and I've certainly not used all of those books in games I've run.
Anyway, that was kind of a distraction from the point of the original post.
Adventure Paths are expensive, risky, and difficult to pull off. They're not something a single person can really do, since the rate at which the parts need to come out to maintain the momentum and excitement outstrips the ability for one person to create, devleop, and edit the thing.
And if you don't start with a bang, you're stuck with a line that's DOA; if you cancel that line and try something new, your customers already are predisposed to not trust you since you didn't follow through on the first one, event if it was a failure from the start.