Only coastal animals would leave a fossil record in Afar^{[1, 2]}, that is, animals living along the coast on south-western Danakil, and whatever lived up in the mountains would be lost (fossils are not as well preserved in those habitats as they are in lacustrine sediments with lots of clay. ) It is well known that geographical isolation causes speciation, and there are few reasons why extant gorillas could not originate from Danakil as well, and have evolved in the mountain regions with few predators, there being few predators on the island as a whole.

In the scenario where extant gorillas originate from Danakil, the geographical proximity with Ardipithecus could have led to interbreeding, between 6–4Ma, and given rise to the Paranthropus lineage with Paranthropus deyiremeda^[3] as the oldest fossil from that lineage, dated at 3.3–3.5Ma. The Paranthropus could further have spread Pthirus gorillae to Australopithecus, and acted as a genetic intermediary between the two species.

Yohannes Haile-Selassie writes in 2015^[3] on the discovery of A. deyiremeda in 2011,

There is now incontrovertible evidence to show that multiple hominins existed contemporaneously in eastern Africa during the Middle Pliocene. Importantly, Woranso–Mille is geographically very close to Hadar (only 35 km north), where Au. afarensis is well documented. In combination, this suggests that multiple hominin species overlapped temporally and lived in close geographic proximity. What remains intriguing, and requires further investigation, is how these taxa are related to each other and to later hominins, and what environmental and ecological factors triggered such diversity.

As the geographical isolation on Danakil ended, the island gorillas would have migrated to the Ethiopian plateau and to increasingly higher elevation where it was safe from predators, while the Autralopithecines and Paranthropus would have migrated along the rift valley, and developed a co-existence in a form of symbiosis, one being brainier and the other more robust, with a more acute sense of smell to discover predators, and with fur that kept them warm, possibly producing heat for the Australopithecines if they slept close together.

In a gorillae-paranthropus-australopithecus scenario, there would have been gene transfer between 6–4 million years ago that is remnant in extant humans, which could explain what is often explained with parallel evolution in a few dozen genes. Subcutaneous fat, for example, as a water adaptation in Australopithecus, could originate from the gorilla lineage, as well as genes for hearing that gorillae acquired to avoid predators while foraging stems and leaves.

The question then is what made the island gorillas more competitive relative to the mainland ones, and why their gene pool (and Pthirus lice) survived better. That could be that genes for brain development^[4] were passed from Australopithecines to gorillae.

References

1. Tectonic history of Danakil and the origin of Pliocene vertebrate fossils in the Afar region (2017)

2. Fossil record of a counter-clockwise rotation of Danakil (2017)

3. New species from Ethiopia further expands Middle Pliocene hominin diversity (2015)

4. Insights into hominid evolution from the gorilla genome sequence (2012)