Interesting project. How big is big? There is a limit to the size of the map, both in game units (distance) and in the number of brushes, but I don't know the numbers by heart. There are some maps that perhaps you don't know that could give you an idea of what is doable:
Dangercity. A classic railing place. It is not terribly big, but at that size and with less bulky buildings you definitely can make a good city maze. The links are for downloading the map--first one--and to a lame youtube video:
http://lvlworld.com/review.php?id=782http://www.youtube.com/watch?v=EBkYyqCyvrc&feature=relatedSome-other-no-name-city. This one is quite impressive. The guy really put some work on the realism side of things (and much better marketing too). There is a download link in the description.
http://www.youtube.com/watch?v=OcAGrF1ZVZMThe Simpsons. Maybe not quite the style you have in mind, but this map is already stretching what is reasonably doable in terms of size and detail. The video is extremely annoying but it does its job:
http://www.youtube.com/watch?v=34LtrnnXQTc&feature=relatedAs for how to proceed with your project, you are probably the only person who can say. It sounds like you need to think carefully about the map layout, which will serve a very particular purpose. You apparently want to do some experimental measurements, so you want the map to have some design features that will make your life easier. So, perhaps, you want it to have some routes easier to follow than others and grade them. How many people got stuck in route #3 as opposed to route #1? How long to complete route #2? Who knows. This could be done playing with the number of dead ends, cross points, z-axis motion, mixture of indoors and outdoors locations, similarity between different places, what have you.
The technical aspects of mapping are quite manageable, especially when you don't have to consider the gaming specifics, i.e. you won't give a monkey about item placement, playability, optimisation (within certain limits, of course) and so on. You know the game, so you already have an idea of what can be done. I suggest you to have a careful thought about what you want (hand made drawings and photographs of real places will help), and then you can worry about how to make it. There is a lot of information regarding mapping, but you may want to come back to ask for the useful ones and for help with particular things you get stuck with. Also, beware that it can take a fair amount of time--months--to achieve the final result, so you should plan ahead and set some goals and priorities; what's the minimum amount of detail you can get away with? how realistic does it need to be for the experiment/project to work?
As a last thing, you will probably want to make a quick and dirty map first, where the important elements are present but with basically zero detail and the absolute minimum lightning work to get you going. Modify it until happy and then proceed with the rest. Apart from saving you a lot of work, you may discover at some point that enough is enough and no more is needed. The workflow for map creation goes like this: geometry-->textures-->lighting. Focus in the geometry the most you can at first and your life will be easier.
Good luck.
PS. Don't forget to show the results!