Some excerpts follow.
But to be honest, as much as I hate calculator use in school, in this age of calculators and computers, efficiency at hand computation is not, IMO, the most critical math skill for kids to learn. I am NOT saying that it should be ignored, or that kids should be allowed to skip it, and just use calculators in class (see rant linked above). But it is not, IMO, the be all and end all of math education, nor is it a prerequisite, IMO, for studying anything else.
What I consider even more important is a strong sense of number. I want kids who know immediately when the answer they got (either by hand computation or with a calculator) is way off. I want kids who have an instinctive understanding of the distributive law before it is ever formally taught or named (12 sevens is obviously the same as 10 sevens and 2 more sevens). I want kids who know when the amount of change handed to them makes no sense. I would rather have a kid who can multiply 64 x 25 mentally (by halving 64 twice and doubling 25 twice, to see that it's equal to 16 x 100 = 1600) than a kid who can sit down and carry out the long multiplication with pencil and paper, by rote.
I feel that we need to consider several things when it comes to calculators. It's best when the kids can do mental calculations and do paper-and-pencil methods, including understanding why they work. Calculators should be used judiciously, but used. Like she mentions, number sense is of paramount importance so that kids can estimate their answers and tell if the calculator "got it wrong" (e.g. they punched wrong buttons).
On spiraling curricula:
Steve is right that a spiral curriculum can lead to a lax attitude of "it's ok if they don't master this now, because they'll see it again later" that goes on ad infinitum, and the kid never masters anything. This is clearly no good. But the solution isn't necessarily to take away the spiraling for those who need it, IMO. The solution is to have limits - for example, it's ok if they don't completely "get" long multiplication when it's previewed in 3rd grade, or even when it's introduced more formally in 4th, but they have to get it when it's reviewed in 5th, or they shouldn't move on.
This sounds like a sane approach.
She also talks about including non-routine problems for ALL students to solve. I definitely recommend this practice and have written about it before! MathMom gives several good reasons for this:
1. First, it provides a fabulous way of helping students to appreciate the uses of the procedures and skills they have learned or are learning.
2. Second, this is the kind of thing that "real mathematicians" do! ... A "mathematician" does not sit down and solve 25 ratio and percent word problems, knowing exactly which skills are required to perform the computations. Instead, she investigates "puzzles", looks for interesting patterns makes new discoveries, generalizes results.
3. Third, it develops self esteem and confidence.
4. Fourth, it builds transferrable problem solving skills.
Read it all at Ramblings of a math mom: Math wars.