What is more interesting, is the attempt by Shah et al. to date the age of the admixture event to ~200 years ago. Note, however, that this estimate was done using ROLLOFF, which produces about half the age as HAPMIX and StepPCO, two other methods of using linkage disequilibrium to date admixture events.
Also, given that ROLLOFF and HAPMIX were done by some of the same people, the discrepancy requires an explanation. The availability of ROLLOFF was mentioned in the Moorjani et al. paper which introduced it, and I am told that a version of it will eventually be released.
It would be a good idea to attempt to explain the ROLLOFF/HAPMIX/StepPCO discrepancy, otherwise, I fear that LD-based age estimation will suffer the fate of Y-STR based age estimation for Y-chromosomes, with two incompatible methods persisting in the literature for years to come.
It seems that there is no shortage of factor of 2+ age discrepancies in the genetics literature no matter where you look, which is often underappreciated by the consumers of the genetic studies.
Anish M. Shah et al. Indian Siddis: African Descendants with Indian Admixture
Ankita Narang et al. Recent Admixture in an Indian Population of African Ancestry Link