perhaps a more useful service would be to send some communications relay and navigation satellites to Mars, along the lines of the Mars Internet idea. I honestly don't know if the Deep Space Network could handle that many rovers along with all the other missions out there. There's also the question of having sufficent ground crew and related facilities to handle all those rovers. The two rovers currently on Mars have stretched the ground crew pretty thin.

A good communications and navigation infrastructure at Mars could pay enormous benefits. For example, the navigation capabilities could allow high precision landings in areas that might otherwise be unreachable. Communications relay capabilities mean each rover and lander wouldn't need to have expensive and relatively heavy high gain antennas and powerful radio transmitters to reach all the way to Earth. Instead, they'd only need to transmit data to orbiting relay satellites that would then send it to Earth, probably using store-dump technology.

Developing these satellites would be far less expensive than sending a bunch of rovers and could do a world of good for Mars exploration. That capability could be useful for the moon, too. You might be able to send 2 or 3 of the satellites on a single Delta II (or Falcon V) launch, further lowering the costs. The multiple satellites could be carried on a common bus that provides all of the needs for the trip to Mars and orbital insertion, then separated once orbit is achieved. There are things you could do to keep the costs reasonable, such as using the TT&C system from the carried satellites to control the bus. The bus itself would be a pretty simple and dumb device.