The parallel occurrence of replicated species-pairs in similar environmental contrasts may arise through a variety of evolutionary mechanisms. In particular, whether parallelism reflects a common history of divergence or repeated parallel divergence driven by divergent selection needs to be ascertained. Reconstructing historical gene flow is therefore of fundamental interest to understand how demography and selection jointly shape the genomic landscape of species divergence. Here, we extend the current modeling framework to explore the multiple facets of speciation-with-gene-flow using demo-genetic divergence models that capture both temporal and chromosomal variation in migration rate and effective population size. We implement this approach to investigate the divergence history of a young adaptive radiation involving five sympatric Lake Whitefish limnetic (dwarf) and benthic (normal) species pairs (Coregonus clupeaformis) characterized by variable degrees of ecological divergence and reproductive isolation. Genome-wide SNPs obtained by RAD-sequencing were used to document the extent of genetic differentiation and of allele sharing among species pairs. Using a composite likelihood approach, we then compared the predictions of 26 divergence models to the unfolded joint allele frequency spectrum of each species-pair. We found strong evidence indicating that a recent (circa 3000-4000 generations) asymmetrical secondary contact between expanding post-glacial populations has accompanied independent Whitefish diversification. Our results suggest that heterogeneous genomic differentiation patterns have emerged through the combined effects of linked selection generating variable rates of lineage sorting across the genome during the allopatric phase, and heterogeneous introgression eroding divergence at different rates across the genome upon secondary contact. This study provides a new retrospective insight into the historical demographic and selective processes that shaped a continuum of divergence associated with ecological speciation. This study also illustrates the efficiency of the newly developed models towards improving the understanding of historical events in the process of adaptive evolution and speciation in any other species complex.