hc44: I'm gonna go against the grain here and say it's a compromise which didn't have to be made. You shouldn't have to re-invent the whole camera because of a sensor change. A sensor can be though of in abstract terms as a plug-able component with a fixed I/O interface. Resolution can be a variable which the rest of the camera can be made to work with as a variable. A hi-res raw image can be down scaled, sometimes you send the hi sometimes the low, you have the option. The engineers shot themselves in the foot when they began the project allowing for a fixed res sensor only, and did so knowing technology advances. In the 60s the Apollo project got a man on the moon in under a decade, imagine how fast pasted their project was, but now they can't change a camera sensor in the space of 8 years.
No, the camera isn't part of rocket science. It's just NASA doing what they do best - going with the easiest solution that has the most data. This is why they are going back to old designs (or at least were, before tons of funding was cut) for new spacecrafts. It is far easier to gain funding and keep a project going that can have funding cut at any second, by showing that you are using proven data or a cheaper way of going about things. NASA engineers are still, at the end of the day, just like everyone else who wants to keep their jobs. I have people who work at NASA in my family and it's just the way it is. Innovation is not really worth much on a spacecraft, but tried-and-true is.