Natsuya points me to this new paper on the distribution and likely spread of Y-DNA haplogroup C, with special focus on East Asia and subhaplogroup C3.
Hua Zhong et al. Global distribution of Y-chromosome haplogroup C reveals the prehistoric migration routes of African exodus and early settlement in East Asia. Journal of Human Genetics 2010. Pay per view (but you can read it freely at ZohoViewer).
CONCLUSIONSNothing too new as you can see, however the the data detail and the neighbor-joining trees of haplotypes for C3 and C5 are of some interest, suggesting that C3 probably coalesced a Mid-East Asia (China and surroundings) rather than SE or NE Asia.
We demonstrated the phylogeographic distribution of one of the most ancient non-African Y-chromosome lineages, from which we inferred the prehistoric migration and expansion of the Hg C lineage. We propose that Hg C was derived from the African exodus and gradually colonized South Asia, Southeast Asia, Oceania and East Asia by a single Paleolithic migration from Africa to Asia and Oceania, which occurred more than 40KYA. The prehistoric northward migration of Hg C in mainland East Asia likely followed the coastline and is consistent with the northward migration of other East Asian Y-chromosome haplogroups.
C6 was not located anywhere, suggesting that it's just a very minor clade. In contrast South Asian specific C5 was detected among 2.5% of Indians and 1.5% of Southern Pakistanis, reminiscent of a back-migration from SE Asia by the same "coastal route" that was used to reach SE Asia at an earlier moment.
It's interesting also the still rather high amount of undefined C* found in various areas: 4.6% in Philippines, 5.6% in East Indonesia, 5.9% in Micronesia, 5.6% in Australian Aborigines, 9.1% among the Mulau of Guangxi, 6.9 among the Shui of Guizhou, 7.6 among the Yao of Guangxi, 7.7 among the Tujia of Hubei, 8.2 among the Hui of Ningxia and 6.7 among the Hezhe of Heilojiang. It is also potentially interesting the 0.5% in India because it might have greater densities in some of its very diverse populations (but it's treated as a single homogenous sample in this paper).