Lactase persistence parallelism

Nicholas Wade has an article in the Times this weekend concerning lactase persistence in Africans, which appears to have arisen in three separate mutations, each different than the mutation leading to lactase persistence in Europeans. All four of these mutations are very recent, and they constitute some of the strongest examples of positive selection observed in human populations:

The principal mutation, found among Nilo-Saharan-speaking ethnic groups of Kenya and Tanzania, arose 2,700 to 6,800 years ago, according to genetic estimates, Dr. Tishkoff's group is to report in the journal Nature Genetics on Monday. This fits well with archaeological evidence suggesting that pastoral peoples from the north reached northern Kenya about 4,500 years ago and southern Kenya and Tanzania 3,300 years ago.
Two other mutations were found, among the Beja people of northeastern Sudan and tribes of the same language family, Afro-Asiatic, in northern Kenya.

The article says "10 times as many descendants", which isn't correct -- the selective advantage of these alleles is only on the order of 10 percent, not 1000 percent!