First, since the overall width of the concrete slab will most likely remain constant the additional weight issue of moving the girders outward is in my mind a non-issue. The only thing that may change would most likely be the rebar requirements for both Positive & Negative Bending Moments + Shear, which may or may not increase either the rebar size or decrease the rebar spacing, not increase the slab thickness. We're probably only looking at a 100 PSF Live Load here for pedestrians. Since there won't be any snowblowing equipment on the bridge due to a lack of sidewalk ramps, this too is a non-issue. Most likely there will be only enclosed stairwell towers or outside stairs, and/or small elevators (for the Handicap assessability laws) provided at each end of the span.
Another thing for our young engineering student to consider in his design would be the use of 4-inch or 6-inch think Prestressed Concrete hollow-core planks in lieu of cast-in-place reinforced concrete slab with shear studs. The underside of the plank ends can be furnished with several embedded steel plates (with industry standard anchorage studs or wire hoops) at each end of the planks to facilitate the welding to planks to girders (at the inside edge of the girder top flanges). The utilization of these planks instead of a cast-in-place slab has several distinct advantages, such as: less weight, significant less cost, and increased speed of construction.
Utilization of these Prestressed concrete planks may actually reduce the size and weight of the two main girders, resulting a goodly amount of principal costs in the project.....less steel material, less steel preparation, less welding, less primer & paint (unless Cor-Ten steel is used), less cast-in-place concrete costs, and less erection costs! A thin (2-inch min.) lightweight latex-modified concrete overlay with weld-wire mesh will need to be installed atop the planks to fill any gaps between them. I envision that the plank erector can get these planks alone laid within a single working day. The provision for the secondary steel framing will still be required for structural considerations such as: bridge stiffness; wind loading; and, lateral stability, particularily at the compression flanges. As an added benefit, the amount of positive camber can be increased due to a lessening of the slab weight (dead load).
Also, provisions for expansion joints would be simplified as well. With this type of slab approach you only need to provide a 1/4-inch thick vertically-oriented closure plate that has been welded to the outer girder top flange edge in the fabricator's shop, thus avoiding expensive and time consuming field welding and increasing QA/QC. It'd also act as a concrete screen when the concrete topping is placed. Also, I strongly suggest that the along the outside edges of the girder top flanges that shear studs (like Nelson studs) be provided in the cast-in-place concrete pocket (located between the continuous steel closure plate and the ends of the planks.....ends of planks must be stuffed with red rosin building paper to prevent concrete disposition inside the open plank cells, unless of course rebar and concrete is to be provided in the ends of the planks for additional plank tie-in requirements). I wouldn't use anything less than a 6-inch wide full slab depth concrete pocket along the outter edge of the top flanges, so as to provide adequate pullout strength of the slab, better bonding around the studs, and lastly it would promote ease of construction. Use of a single row of these shear studs will undoubtedly NOT help with regard to composite action, but rather they will promote tying together of the various structural elements.
In regard to compression or tension of the cantilevered concrete slab segments. The top surface of the slab within the cantilevered portion of the slab will be in tension whereas the bottom portion of the slab will be in compression. If there is any sort of "Fixity" of the slab at it's supports, then even the portion of the slab spanning between the girders will have negative bending moments (tension in the upper portion of the slab, necessitating rebar installation in the upper region of the slab) occurring there, located between the girder centerlines to the Design Inflection Points (IP) that'd be so located somewhere around both 1/4 span locations, give or take. The remaining portion of the slab span between the IP's will fall within the Positive Bending Moment realm. And since the overall span of this slab may be somewhere between 12-to-15 feet (the OP hasn't given us this information. or little else for that matter), it may be best to provide upper and lower steel reinforcement (ie, Doubly-Reinforced Slab) throughout the slab span transversely (one-way slab action). Additionally, longitudinal reinforcement must be provided in the slab as well for concrete Shrinkage and Temperature Steel requirements.
I hope this helps clear up some the questions, as well as providing design alternatives.
Have a great sunny day people!
"Veni, Vidi, Vici"; hendiatris attributed to Gaius Julius Caesar, 47 B.C.