Both the original paper and the two letters use the "evolutionary" mutation rate, and hence the given dates should be divided by approximately 3 to obtain realistic dates.
Excerpts from the two letters:
On the origin of Y-chromosome haplogroup N1b
Boris Malyarchuk and Miroslava Derenko
Eur J Hum Genet advance online publication, June 17, 2009; doi:10.1038/ejhg.2009.100
A possible scenario explaining the phylogeographic pattern of hg N1b is that the subclusters N1b-A and N1b-E might have diverged from a common ancestor
anywhere in Siberia approximately 15 ky ago (with upper bound of divergence time, TD, between these subclusters estimated (according to Zhivotovsky9) as 14.7±5.9 ky). According to STR variation, expansion of Northwest Siberian/Uralic/Northeast European branch, N1b-E, may have occurred approximately 7 kya and the expansion of South Siberian branch N1b-A about 5 kya. Divergence time between South Siberian and Northwest Siberian/Uralic N1b-A haplotypes is 4 ky (TD¼4.0±1.0 ky), so this suggests that expansion of N1b-A lineages in Northwest Siberia and Urals
(about 3 kya) might have occurred right after their migrations from South Siberia. Future studies of Y-chromosome variation using a combined STR-SNP approach may enable a better definition of geographic origin of N1b-P43 mutation.
Reply to B Malyarchuk and M Derenko: a need for further investigation of Uralic and Siberian populations in the search for haplogroup N1b's origins
Sheyla Mirabal, Peter A Underhill and Rene J Herrera
Eur J Hum Genet advance online publication, June 17, 2009; doi:10.1038/ejhg.2009.101
The higher variance and age estimates (Tables 1 and 2) for both European and Uralic populations as compared to the Siberian groups lend support to our previous statement suggesting that N1b may have arisen in the Uralic range or in an area/region proximal to it; however, a much more exhaustive investigative effort must be conducted to trace the origins of N1b and its branches N1b-A and N1b-E.