Well, although the original poster is from 2011, it seems there were some more comments last months so I'll add my 2 cents.
Without having gone to
CMU, I'd say go to
CMU if you are worried about landing a job. It is a very well respected university all around in addition to having a top mfe program. Also, if you're ending up paying roughly the same tuition in either case,
CMU seems like the best bet.
That said, the Rutgers MSMF program is not at all bad. Especially if you get in-state tuition. Especially Especially if you actually *want* to live in NJ

. I am in MS Stats in Rutgers and MSMF at Rutgers (getting two MS degrees was overkill in hindsight, but live and learn) and I am almost finished with both degrees (8 courses / 10 MSMF, 9/10 MS Stats).
I have taken all the Stats courses that overlap with MSMF (there are about 4 such courses of which you can use up to 3 for both programs ; out of 10 courses required for MS Stats and 10 courses required for MSMF). 30% of the courses I guess can seem like you should have a handle on what the MSMF is like, but it's really not a representative sample.
The way the MSMF works is that some of it's courses are offered in the Applied Math department, some are offered in the Statistics department, some are offered in the Electrical Engineering Department and you can ask to have approved for electives relevant graduate courses from departments such as Operations Research, Economics, Finance, Computer Science, etc. It is open to the criticism of being called a 'collection of courses'.
However, the core of the program is the 5 or 6 Applied Math courses that are attended almost uniformly by MSMF students (and the occasional curious Math PhD student) which are 2 semesters Derivatives Pricing Mathematics, 2 semesters Numerical Analysis, 1+ semester <pick one or more>{Portfolio Theory, Credit Risk Modeling, Computational Finance}.
I personally don't have a problem with the MS Stats courses I took (though I had to maneuver things carefully to ensure that most of the sections I was taking were being taught by Rutgers academic staff, not contractors ; I've never had to do this for any department in RU except Stats). The core of the MSMF courses listed above are much more rigorous than the MS Stats courses (as I mentioned I have taken many courses in both categories) and they are qualitatively different in that the focus is on mathematics not statistics.
If I were to tell someone about the Rutgers program (i.e. someone who didn't have an admit from
CMU), I would say my experience is generally positive but note that:
1) Rutgers program is on the academic side, some days I love this and some days it drives me crazy, but at the end of the day you'll spend more time writing mathematics than doing 'gee whiz' type projects (there's plenty of opportunity to code within the program however);
2) In the same vein, Rutgers tries to balance theory and applications and is currently a bit heavier on theory side IMHO, again it's good and bad;
3) Rutgers MSMF does not teach much traditional Finance. There are electives but they are not that convenient to take (you'd have to travel a bit for them from the main Busch campus). I personally don't see this as too big a deal. The justification that this is the sort of thing you could learn on your own versus the opportunity cost of not taking a course from one of RU's well respected quantitative departments (see above) makes sense to me now at least (we'll see what I think when I apply for jobs!);
4) Live on campus if possible. I live with my family within a commutable distance to Rutgers but getting around the actual campus in a car is a pain in the ass.