This is also a good way of looking at it. VentureStar is probably a bad example because it is unclear it would have worked at all and even if it did it was far from where Phoenix is now.

Also to comment about "using things that flew before" is often a problem because it is not available. Remember MER/MPL were started over 10 years ago and were designed with reliable technology then. Companies stop making parts after that long (ask Intel for Pentium 200 MMX or a 486) and this is even more true when it involves space qualified parts, which involve a huge amount of paperwork and testing.

It would take a lot to cancel Phoenix because of a budget overrun. Remember that hundreds of millions of dollars have already been spent on the project and to cancel it would waste everything. The DAWN project had a similar review and was put on hold for months but was eventually reinstated.

Cost capped missions are both good and bad. It has made NASA pay more attention to cost but one of the problems is that NASA still likes to pick aggressive - high value missions. Deep Impact, Phoenix, Messenger, Dawn are all not simple - MGS like missions (i.e. orbiter). This rewards proposers who make "aggressive" assumptions about what a mission will cost... which then leads to overruns.

While MCO failed due to the infamous english to metric mix up one of the core issues that contributed to the failure was lack of funding. When you lack funding, you lack people, and then things start slipping through the cracks. Who is to say if MCO or even MPL had more appropriate funding if they would have failed. Perhaps testing or analysis that they probably had to cut for budgetary reasons would have found the problems that caused their loss.

The core problem is that the funding available for cost capped missions does not allow for most projects if everyone was sufficiently conservative with their cost estimates so as to not overrun. The projects that it does allow are often of lower scientific value than the more expensive ones... and the selection is heavily based on science.

NASA might be learning it's lesson however. If you notice the Scout proposals selected to go to the next round do not include any landers.

All in all it is pointless to can a mission that is in ATLO because of a cost overrun. The money has already been spent. If cost is a concern... do not select the risky missions in the first place.

MSL is an RTG powered rover.

Keep in mind there are two other benefits from RTG's.

1. Lots of waste heat which can be used to heat the rover. This is a significant power savings.

2. No batteries.. = mass savings ;)

Somethings to keep in mind...

A camera (any camera), especially a high-end one like HiRISE, will be very difficult to deploy in the Jovian environment. The radiation would make much more difficult to build, and HiRISE was not exactly easy (or cheap ) to build in the first place.

Why a camera? What will you learn? Instruments are deployed on spacecraft to answer scientific questions. What questions will a camera answer?

The ESA mission study is pretty aggressive. The level of radiation shielding they assume for the instruments is paltry compared to the 5mrad assumption of radiation dosage. It is difficult to make digital electronics that can take 50-100 krad... let alone 5mrad with minimal shielding.

I also find the masses/power of the instruments to be very suspect... I usually check http://nssdc.gsfc.nasa.gov/ and look for similar instruments. You will probably be able to build lighter more capable instruments in the future... but not by factors of 2 or more... especially since you will have to use older, lower performing parts to withstand the radiation.