Kids are ready for formal operations when their earlier experience has prepared them for formal operations. In some places around the world, students in fifth grade--including below-average fifth-graders--are already learning algebra. In quite a few whole countries around the world (for example, Taiwan, where my wife grew up when Taiwan was still a developing country) all seventh graders study algebra and geometry as part of their regular school math lessons, yes, including the below-average seventh graders. (In Taiwan, they also learned enough of the International Phonetic Alphabet to transcribe General American English at that age, but now that skill, which most United States reading teachers lack entirely even with a college degree, has moved into the elementary school curriculum.) Good education at the beginning produces better results in the middle grades.[1]
"Learning algebra" doesn't necessarily imply understanding the abstract concepts of algebra. Without the developmental substrate, children instead lean mostly on rote memorization. This seems to "work" when the class is taught in terms of rote application of memorized formulae, but will fall over the moment that students are presented with a problem in a novel context, or expected to synthesize multiple component skills into a larger solution.
If you've ever tried to tutor someone who "struggles with word-problems", the majority of the time it's because they don't actually understand the things they've been "learning" at all.
Similarly, this tends to be the difference between people who "get" programming and people who don't. If you ever talked to someone who failed a programming course, it's usually clear from their approach that every new language or API they learned became, to them, another set of rote facts to burn into their mind, never coming together to paint a picture of what the formal act of "programming" is generally.
[1] http://www.amazon.com/Knowing-Teaching-Elementary-Mathematic...
http://condor.depaul.edu/sepp/mat660/Askey.pdf
http://www.ams.org/notices/199908/rev-howe.pdf