If you think in terms of ice-age climate, with sea-level about 150 ft lower than at present and the Mediterranean regularly covered by thick arctic-like ice in winter, it is easy to imagine early humans making their way back and forth over an ice-covered strait of Gibraltar or along an ice-free coastal strip connecting western Europe with West Africa. I think the discovery of relatively large number of neanderthal genes in West African tribes like the Yoruba is one of those unexpected and unpredicted facts that on further reflection makes a lot of sense, justifying the statistical analysis used. After all, if a statistical algorithm only shows what's expected, you have to wonder whether all it's done is to give a statistical excuse for what's already believed to be true.
Indeed, I think this is a possible explanation. On the other hand, there is just as much danger of post hoc generalization the other way!
Testing that hypothesis will require some more sophisticated estimates of the ages of particular gene regions that have been inherited from Neandertals in West African populations.