According to a new preprint by Durvasula and Sankararaman (D+S):
Using this method, we find that ~7.97±0.6% of the genetic ancestry from the West African Yoruba population traces its origin to an unidentified, archaic population
This ~8% matches well the ~9% of "West Africa A" in Yoruba of the model of Skoglund et al. Figure 3D. If "West Africa A" corresponds to the Archaic Ghost of D+S, then the Mende have the most of it at ~13%.

I have long maintained that the higher genetic diversity of extant Sub-Saharan Africans is the result of admixture between "Afrasians" (a population that spawned Eurasians and much of the ancestry of Sub-Saharans and which had "low" (Eurasian-level) of genetic diversity) and multiple layers of "Palaeoafricans". It would seem that one such layer has now been discovered.

Where did the Afrasians live? Recent developments pushed back the presence of modern humans in both North Africa and the Middle East, making both regions highly competitive as the cradle of the Afrasians. The odds for Sub-Saharan Africa have greatly diminished also by the discovery of late non-sapiens H. naledi in South Africa (which was naively postulated as a cradle based on the presence there today of genetically diverse San Bushmen, but who are not descendants of even Late Pleistocene South Africans), as well as of the archaic component in the genomes of West Africans. These discoveries pile up on top of known archaic skulls of late provenance in both Central and West Africa.

Remember though, that the archaic admixture in West Africans is "less archaic" (more closely related to H. sapiens) than the Neandertal/Denisovan ancestry which contributed to extant Eurasians. All Africans (modern or archaic) are a branch within the phylogeny of Eurasians, with Australoids (and now apparently East Asians too) having the deepest known strain of human ancestry inherited from the elusive Denisovans.