It seems like you should test out your ideas on one city (which is much easier to market), then once you find a successful model, start branching out into other cities.
You could spend some money on PPC ads on keywords such as "San Francisco tours", "San Francisco golden gate bridge", etc, and test out different landing pages and different types of guides.
I also think something like this would work better if the information was all free, and people get to rate the quality of each tour. The ones with higher ratings will be shown higher up - maybe offer prizes to authors who have highly rated tours.