This seems from the reports I've read to be pretty spot on. I would add an addendum to an earlier comment about this being why no nuclear plants will ever be built in the US again though; the current designs are generally "passive fail", meaning that unless electricity is supplied to the control systems, the plant will just... stop being just sub-critical and will go non-critical very quickly. For instance the pebble bed designs. My (somewhat, I'm probably giving myself a little too little credit) understanding is that these plants use nuclear fuel that just... can't react on it's own due to the sheathing materials. I think those are pyrolytic carbon still though, so of course there will still be problems with burning if they are exposed to air, the accompanying release of hydrogen, etc (I think).
This seems very honestly to be the entire focus of the nuclear industry -- designing plants which are safe to operate no matter what, which maintain reasonable cost-effectiveness. It's basically the holy grail.
I think the current problem is:
1. Natural gas is cheap, coal is cheap, they are cheaper to build and easier to maintain.
2. The regulatory process and validation work to get a new plant design is intimidating. Probably even intimidating as compared to the design of fighter jets.
3. Nuclear *is* scary to the vast majority of people. This is residual in large part from Long Island, and based in concerns over running reactors commissioned in the 60s still being operated. *That* part I am scared of. But as a scientist and engineer, I think that these are solvable problems so long as safety and the concepts of "fail safe" systems engineering based on the Therac-25 (http://en.wikipedia.org/wiki/Therac-25) which seem to have very permanently changed the way that people fundamentally think about how to do system engineering. These problems had not arisen and become understood when those plants went into operation. A current plant definitely would do a far better job of that.
Heck, it even effects me on a daily basis (at this point in my career I would classify myself as a systems engineer); I think all the time "What happens if all this equipment just stops working" and the answer is always "go to a safe operational mode". The are different ways to do that. You have the F-16 style of doing that, which includes crazy amounts of unstable control algorithms. But by *far* the preferred mechanism is physical. For instance, if I have a furnace I expect to go to 2000C, and monitor the temperature with one thermocouple while I use a single additional thermocouple as a safety, is not really enough. I would *far* rather have a thermal fuse that blows hard when a temperature exceeds some set ultimate super failure limit and shuts everything off immediately. I don't trust thermocouples to be reliable, and I don't trust the controls equipment to respond properly in an emergency.
But in one of these pebble beds, the sorts of controls they are integrating are way beyond "having power", by far the best safety integration is to design it such that electricity failing causes large physical things to happen. Dumping the pebble bed entirely, or dumping immediately a mediator into the reactor that is only prevented from triggering by constant electricity. Some of the designs I've seen literally place the reactor under a ridiculously large tank of water held closed by electricity. I don't know in what way that would fail, but it would be far superior to what happened in fukoshima.