There's very little data, which is the main problem. In terms of books about KIPP, the one I've read is Whatever It Takes, which is about Geoffrey Canada, who founded the Harlem Children's Zone. I'm not sure if this school is technically affiliated with KIPP proper, but it used the KIPP model.
It's very difficult to explain the problems with KIPP. Basically you need to take a really good look at how the kids are actually spending their days, and then compare it with what we know about child development, educational research, best practices in various fields, research data on student achievement, etc. I don't think there is any way to understand it without having a solid background in education theory, education research, cognitive development, psychology, early childhood education, history, and even things like organization behavior and management theory, etc. The problem of what makes a good school and a good classroom is insanely complicated, which is why we've been using basically the same model since WWI.
Forget the theory. The theory might tell me why some measurable effect occurs and generate a complicated model of how it occurs. But without delving into theory, you can certainly tell me what the measurable effect will be. E.g., "KIPP students will have 10% lower math scores in year 2-4 of college," or something like that.
There isn't enough data to say.
"But without delving into theory, you can certainly tell me what the measurable effect will be."
Cognitively, I'd be most concerned about whether the program would be harming their executive function and their intrinsic motivation. I don't see any value of trying to predict specific metrics though, beyond identifying specific metrics to look at once more data exists.
EDIT: just read http://en.wikipedia.org/wiki/Knowledge_Is_Power_Program#Crit..., which is certainly interesting.