There are any number of types of extended missions you can design for an Oppy whose IDD is no longer available. Some of them might cost less than others.
For example, I draw y'all's attention to the fact that the MERs are capable of autonavigation. One extended mission I can imagine would have us programming Oppy to drive for five days, taking a Navcam pan at every daily stop, and dumping the Navcams and all the engineering data for the past five days at the end of the fifth day. The MER team then analyzes the data for problems (sand traps, failed wheels, etc.) and, if there are no engineering issues to address, decide if they want to spend some time on Pancam shots of interesting items, and make appropriate changes to the upcoming standardized five-day program. I figure they can do a decent job of that, most times, in 2 days or so.
If there's a serious hardware issue on such an extended mission, you're probably looking at end-of-mission or once-a-month-check-in-stationary-rover mission, so you wouldn't run the risk (or expense) of working tons of hours and spending tons of Mars Yard time trying to work around a major glitch.
Seems like that would save some money -- especially the concept of standardizing a five-day driving program that Oppy could execute without supervision. You build it with easy-to-change parameters for things like daily heading and such, so you can adjust your route as you go along.
Granted, you do face the possibility of Oppy running into trouble while no one is looking. But you're only going to ever be, at the most, five days late in seeing a problem develop, and you simply program Oppy to safe herself if something really serious happens.
You'd still need a driving team and a science team, of course, but many of them would be working part-time on Oppy, the rest of their time could be charged to other programs. And you'd cut the number of DSN sessions down by a half, probably, attacking the extended mission costs at their most expensive spot.
The whole idea would be to move as far as possible, and see as many different aspects of the terrain as possible, documenting the scene every 20 to 100 meters. I can think of a time when you could argue strenuously for the value of a Mars rover that *only* did that!
-the other Doug