One collateral issue to the question of Afro-Asiatic origins discussed in two previous posts today (FWIW, those two posts and this one were mostly written about six months ago, but were awaiting final polishing and so didn't get posted until today), is the interrelationship between that question and the Afro-Dravidian hypothesis.
The Afro-Dravidian hypothesis argues that the Dravidian language is a derivative of an African language from the Niger-Congo family of languages, and that much of the cultural and technological packet associated with Dravidian culture has African origins. Frenchman Bernard Sargent is one of its principle advocates. I'm not entirely convinced of this hypotheis, but there is enough data to back it up to warrant giving the idea serious consideration, so I don't dismiss it out of hand either, as many do.
The linguistic evidence from within the Dravidian languages points to an origin of the Dravidian language of India around the time of the South Asian Neolithic ca. 2500 BCE. A few linguists have noted similarities between some Niger-Congo languages (particularly those on the Afro-Asiatic/Niger-Congo boundary area) and the proto-Dravidian language.
The crops used in early Dravidian agriculture were domesticated in and had their origins the African Sahel.
There are no meaningful African mtDNA traces in Dravidian South Asia that can't be attributed to events in the historic era, however, and the strongest outside Y-DNA signal, which coincides quite well with the locations of the proto-Dravidan society are rich in Y-DNA haplogroup T.
Given the fact that ancient peoples rich in Y-DNA haplogroup T were sailing the Red Sea and Indian Ocean from the Horn of Africa at the time, it makes sense that a male dominated group of people from the Horn of Africa might arrived on the east coast of India and bring new crops and technologies around 2500 BCE. Recent genetic profiles of tribal populations in India also support the inferrence that some of these populations may have genetic origins outside India in a time frame similar to that of the arrival of Indo-European and Austroasiatic food production and archeological cultures to India, rather than with the hunter-gatherers of India's deep indigeneous past. So, the fact that some of the high frequency Y-DNA haplogroup T populations in Indian are tribal populations doesn't necessarily contradict an Afro-Dravidian hypothesis.
The crops, technologies and potential linguistic links seem like a better fit to a Sahel agriculture, Niger-Congo language speaking people, however, than to a society that one might expect to be a Cushitic language speaking people then and there in what might have been the Kingdom of Punt or the Kingdom of Cush, which Y-DNA haplogroup T is currently common, and haplogroup T arguably looks like it has Egyptian or Mesopotamian origins, not Sahel African origins.
Why would people with overwhelming Egyptian/Mesopotamian patrilines speak a Niger-Congo language or be familiar with Sahel agriculture? Was Cushitic limited to areas further north at the time, and were Niger-Congo languages (driven by the expansion of Sahel agriculture) a layer present before Afro-Asiatic languages were in the Horn of Africa? Did ethnic Egyptians precede the Cushitic language in the sea trade of the Horn of Africa?
The "substrate" genetics of East Africa after one removes markers that look like Eurasian back migrations and also removes markers strongly associated with one or another Afro-Asiatic linguistic family, in uniparental and autosomal genetics, are distinctly East African and do not suggest prior genetic unity with West African populations prior to a very remote date (ca. 30,000+ years).
Another way to think about that fact is that East Africa has experienced at least two wave of Eurasian back migration. Sometime in the early Holocene (a period that bridges the Epipaleolithic and the Neolithic eras) associated with mtDNA haplogroups like M1 and U6. The other with the arrival of the Ethiosemitic languages. One plausible way to understand the spread of the Afro-Asiatic languages is to guess that it occured at the time of the early Holocene event. But, East Africa has a distinct genetic identity that predates either of these events and this identity is distinct from West Africa. West African sourced African genetics don't appear in East Africa until around the time of Bantu expansion into East Africa, contradicting the plausible thought that West Africans could have had notable demic influences on East Africa starting when the West Africans developed Sahel agriculture. (As Jared Diamond explains in "Guns, Germs and Steel", neither Sahel crops nor Fertile Crescent crops do well in the other's climate due to their differing seasonal patterns, although Sahel crops can do well in the monsoon climate of Southern India which has a seasonal pattern similar to the African Sahel.)
One possibility is that Horn of Africa people who could have brought the cultural package of Sahel agriculture to India may have themselves been recent recipients of that cultural package and experienced language shift in an early wave of Bantu, or pre-Bantu, expansion from West Africa, that was limited to a cultural/political elite and that the Niger-Congo language may have experienced similar linguistic trends to other languages on the Niger-Congo/Afro-Asiatic linguistic border like Swahili and some of the Niger-Congo languages spoken in the general vicinity of Senegal.