As a current first-year Harris student, I figured I'd give my two cents.
To letsjustgo:
1) I don't know of anyone who works full-time while in the Harris MPP program. There simply wouldn't be enough time (at least not for us mere mortals. perhaps it's been done, but I haven't heard of it). Part-time jobs are common, however, particularly after the first couple of quarters.
2) I'd estimate that about 2/3 of the students live in Hyde Park, while the rest live mostly near downtown or on the north side. Keep in mind that unless you have a car (and most students do not), commuting from the north side is quite time consuming. I live downtown (in the South Loop), and my commute by bus is quite reasonable (under 1/2 hour).
3) I think the program is relatively in line with my expectations, so this doesn't apply to me so well. However, I'd say that in general it's important that you enjoy thinking. I hear complaints from time to time regarding the "theoretical" nature of the Harris program. Yes, most of the first two quarters is focused on theory, but if you don't understand basic economic and statistical theory, how on earth do you expect to rigorously analyze policy? The standard core sequences are certainly not unreasonably demanding, so don't let this dissuade you from coming to Harris. If you enjoy theory, and want to stretch yourself mathematically (and have the background to do so), then you can take the advanced econ or statistics sequences. Certainly, these classes are more theoretical, but if that bothers you, don't take them. That's one of the things that I enjoy about Harris; I know that there are advanced classes that I could take (whether at Harris, or the Econ department, etc.) that would never fail to challenge me. It's simply a matter of how technically-minded you are, and how much you can handle. How challenging the program becomes for you is completely within your control, since the curriculum is so flexible after the first two quarters.
To enofotan:
Yes, MPP students can take Ph.D level classes at Harris, and some do. As I mentioned above, it's often a matter of whether your math background is strong enough to allow you to get the most out of these classes. The 441/442 micro sequence assumes familiarity with multivariate calculus and linear algebra, so if you've only taken math through calc 1, I would not recommend taking these classes (started in this sequence, but dropped it since I didn't have a background in linear algebra, and didn't want to overload myself since I was taking the advanced stats sequence).
The 312/313 advanced stats sequence (or "math stats" as it is generally called) assumes familiarity with differential and integral calculus, sometimes in multivariate settings. However, the amount of integrals that you'll have to crunch through by hand is pretty limited; you need to understand them, but you're not wasting all of your time doing these calculations. Experience with linear algebra is not presumed with this sequence. If you'd like something more advanced than this, you can take classes in the statistics department that would better prepare you for the Ph.D level econometrics courses. But again, these require more of a math background.
The majority of students (2/3 to 3/4) don't take either of these advanced sequences, so don't feel compelled to do so. It's more important to make sure that you're placed into classes where you can understand the material without constantly struggling with the math involved.
Hope this helps!