When the results of a collaborative project between MIT and Stellenbosch studying the population genetics of a unique South African racial group shows up on PubMed, I really have no choice but to give it some attention.
First set of background info: in South Africa, there is a group of people known as "Coloreds" who make up about 9% of the total population, but in Cape Town, Coloreds make up a whopping 48% of the population. They are by far the largest demographic group here - second place goes to Blacks with 31%, and third is won by Whites, with 19%. The Coloreds are the result of hundreds of years of interbreeding between the whites who lived here (mostly Afrikaners), their slaves, and other random Africans who just happened to be hanging around. During Apartheid, the Coloreds were given more rights than the Blacks, but their lives were still made to suck.
So you've got this odd ethnic group which has developed over the past half millennium. I already described which groups historians think contributed to the Colored ethnicity, but it isn't a very satisfactory explanation when one wishes to ask such question as to just how much each progenitor race contributed to the formation of the Colored ethnicity.
Since creating a new race is not a particularly common phenomenon, it is an interesting thing to study. Understanding the genetic breakdown of colored South African DNA would arguably be helpful for understanding their medical conditions. Also, I think its a great case study which will become more and more relevant as mixed race breeding grows.
This was a descriptive study, and I like those because unless they're just flat out wrong (unlikely) they're just observations about how things are that we didn't know yet. All around useful stuff. Now that we know that the main contributors to the Colored race are (very very likely) Europeans, South Asians, Indonesians, and IsiXhosa relatives, one can make use of this information for medical purposes, cultural investigations, further racial foundation analyzes; all kinds of stuff.
Unless the methodology is severely flawed, and I haven't taken the time or obtained the expertise to check, then this is a solid paper.
Thanks to Dienekes over at his own genetics and anthropology blog for the link on this one.