Landers again

At risk of beating this topic to death...
This write-up is quite a bit different from my earlier Ceres proposal, but I think it would achieve the same objectives.

 NASA stands to gain significantly in the upcoming omnibus spending bill, as reported by SpaceNews. Space technology continues to get the shaft, while SLS and planetary science get boosts. Overall the agency is set to receive $756 million more than requested, for an overall award of ~$19.3 billion. Much of that is earmarked for specific projects. Most specifically, $175 million is earmarked for design of a Europa mission with a lander.

 Let's assume that this modified Europa mission will be roughly New Frontiers-class ($750 million to $1 billion). The design and development costs for the program will be finely-tuned to this one mission's requirements and eat more than two-thirds of the cash. The one-off flight hardware will be quite expensive. It will probably meet its primary and secondary objectives and return a lot of good science data.

 Instead of following this process, what would happen if we handle the design and development as a generic outer-bodies exploration vehicle program leveraging experience from the several successful deep space probes NASA has already fielded, most recently New Horizons?
 Consider the MCSB (modular common spacecraft bus) as a model program but scale it into the 20-40 ton fueled mass range, sized for an Ariane, Delta IV Heavy, Proton, Vulcan or Falcon Heavy main payload or as an add-on payload to an SLS launch.



 In particular, this would become the standard heavy bus for future exploration missions, with a handful of standardized modular options. This effort would be slightly more expensive than a Discovery-class mission ($450 million cap vs. ~$500-600 million program cost), but the deliverable would be a fleet of exploration spacecraft for use in follow-on programs.
 Let's suppose after margins, scope creep and cost overruns that the design effort costs $700 million and produces three flight-qualified craft. This seems like a very high estimate given the MCSB's cost of about $84 million, but this is a large and complex bus with expensive testing requirements. $700 million is in line with 2/3 of a NF-class mission plus overruns. It also gives NASA room to maneuver with contractors if necessary, perhaps even carrying two designs through to the prototype phase.

 Each spacecraft would consist of an orbiter segment (main propulsion, navigation, communications relay, imaging, other orbital instruments and a cubesat dispenser) and a lander segment (landing hardware/software, ground instruments, rovers, other payloads). The actual instruments would not be included as these would depend on the target body and desired science return. In most cases the orbiter section could be retasked with extended mission objectives depending on onboard resources. The bus would be designed to handle either chemical or ion main propulsion, with the ion power management hardware as a modular plugin separate from main bus power.
 Future craft could be produced for $100 million or less, with essentially no additional software or hardware development for new craft. Missions without a landed payload could simply omit the lander segment or swap it for additional cubesat dispensers or fuel. Incremental improvements like better solar cells could be incorporated with minimal changes. For reference, this is less than an F-35 fighter jet on average; we plan to buy some 2,400 of these over that craft's life. Diverting even 1% of that program's flight hardware costs would be enough to put a lander on every major moon in our solar system plus a handful of comets and asteroids. Add in the 1% maintenance and operations costs that would be saved and we could afford to capture four asteroids a year ARM-style for the next decade in addition to landing on the major moons. Imagine what we could do with 20,000 tons of asteroids in a convenient location.

Back to the mission. Now we have three flight-worthy craft available, sans instruments. Instrument and launch costs plus a modest amount for operations are still needed.
 - One goes to Luna as a shakedown mission, perhaps carrying a sample return payload to Shackleton Crater. Extended mission option might be an EML2 deployment as a telescope. Cost is about $70 million plus payload and can (should) be executed as soon as the first flight hardware is available.
 - The second heads to Ceres for low-gravity icy body operational testing, plus the usual variety of surface and orbital instruments. Extended mission option would be as an asteroid belt telescope. Cost is closer to $100 million due to pricier instruments and may include refinements from the lunar test.
 - The third heads to Europa with enhanced radiation shielding. Extended mission is unlikely due to the incredibly hostile environment. Cost is perhaps $120 million and may include refinements from the Ceres test. An ion-powered cruise stage would be more expensive to build and operate, but would be much more capable; the trade-off depends partly on the launcher's capabilities. If an SLS launch is required this will be more like $900 million.

 The overall program ends up costing about $990 million, near the upper end of what the Europa mission would have cost. That assumes a commercial launch for the Europa payload; an SLS launch would nearly double the cost of the project. With overages, mission extensions and unforeseen events it could run up to perhaps $1.2 billion. We would get science return from three locations rather than one (though perhaps less comprehensive data from Europa than originally planned). This would include data from identical instruments on three different worlds. We would get a standard bus for further exploration missions; that alone could save $100-200 million in design costs (let alone reduced assembly costs) per Discovery-class mission and pay for the development effort in under a decade.
 Having a proven bus, software package and standardized instrument and payload interfaces would turn the scientific exploration craft into a commodity. NASA could afford to keep one on hand as a flight-ready engineering reference, available to be launched on short notice. Contractors would be able to maintain their construction capacity through frequent purchases. It could even become feasible for entities without a major space program to purchase instruments, the bus and a commercial launch and do worthwhile science without requiring massive infrastructure investment. This is already possible for communication satellites; getting a pricetag of $200-300 million for an interplanetary exploration mission would open doors in the same way that the commercial commsat market did for broadcast.