You couldn't use aerobraking to put the vehicle into a geosynch orbit around Mars. You'd need to go back to the tried and true (but expensive) technique of slowing to orbital velocity using thrusters, then circularizing the orbit using more thruster firings. It could certainly be done, it'd just cost quite a bit to use all of that fuel.

If they're going to do this, I'd like to get the most bang for the buck possible. I'd combine the GOES type sensors with a data relay capability, perhaps with some navigation signals thrown in for good measure. This would be a grown up version of the Mars Network concept. You'd need 3 satellites to get global coverage of the surface plus some smaller microsats in inclined orbits to supplement the navigation signals and fill in any communications gaps. This idea would complement Mars Network, not replace it.

Doing it this way would accomplish several purposes. First, you'd get those GOES-style images of Martian weather. Second, you'd provide a high bandwidth communications infrastructure that a host of unmanned and manned vehicles could use. Third, the navigation capability would be useful for rovers and precision landings.