We know the fossil record underestimates diversity at least a little, and we know that forested environments in Africa tend to be underrepresented. Given this, the diversity of Miocene apes may have been rather impressive, because there is a fairly high diversity in what we can assume is a biased record.
But I’d like to make the argument from another angle, that of modern ecological analogues. Let us assume that the greater apparent diversity of apes in the middle and late Miocene compared today can be accurately translated as a modern reduction in ape diversity. Not counting the relatively diverse lesser apes, there are five species (2 chimps, gorilla, human, orang) which can be further divided into 10 subspecies, across the entire old world.
Now look at the size range of all of the living apes. Gibbons are the smallest and gorillas the largest. When a family or subfamily of land mammal is diverse in a particular region (a biome or something larger than a biome) we tend to see that diversity played out along a spectrum of size, and against size we can find additional diversity derived from dietary or subhabitat differences and geography. It seems to me that there is room in the size spectrum between gibbons and chimps, and orangs and gorillas, and there is certainly room above the gorilla size as indicated by the existence in the fossil record of very large Asian forms.
We know that some of the later Miocene apes were bipedal, and it is starting to look like bipedalism or something like bipedalism is showing up among other apes in the Miocene as well. So perhaps there is a spectrum of locomotory pattern along which diversity may be spread.
This gives us a the following size classes: gibbon, siamang, [something in between], chimp, orang, [something in between], goriilla, [something bigger], or at total (a minimum?) of eight size classes across which apes might exist in a world in which apes are divers. Like the Miocene. If we add to this a more arboral form and a more bipedal form, perhaps we double the number, or perhaps we add about five new classes (I’m guessing that a Mighty Joe Young size ape would not have been bipedal!). This gives us about a dozen, conservatively estimated, niches when we divvy up size and so-called positional behavior.
To this we can add geography. It is probably reasonable to assume that a wetter, more forested middle and late Miocene Africa could be divided into at least four or regions, between the West/Central divide that modern biogeogrpahy tells us was effective at least in the Late Miocene, the Congo River divide, North/Central Africa, East Africa and Southern Africa. Let’s conservatively assume four, and let’s assume that only half (six) of the hypothesized ape species are divided among these areas. That means that 24 species are endemic to varoius regions, and six additional species are more widely spread for a conservative estimate of 30 species.
Among these species there may have been several bipedal forms, but only one of them (plus or minus a little hybridization hanky panky here and there) would have been the human ancestor. Of course, no one at the time suspected that …. (Or they probably would have done something about it.)
This is not an outrageous suggestion. The idea that if you went back in time to a more ape-rich time (and we know it was more ape-rich) and got a current copy of the Guide to the Mammals of Africa, the ape section would have a few dozen species, just like the monkey section or the antelope section today has a few dozen species.