My theory (the evidence is s solid that it is well beyond a hypothesis) is that Eurasia was repopulated from Australia around 60kya, but I do not say it is impossible some Africans emerged into Eurasia in the period 60 - 50Kya. They do not seem to have based on existing data.
There is high genetic diversity in Sub-Saharan Africa, yes, but also incredible diversity in Australia. Right now nobody can state that the African diversity is higher than that of Aboriginals. Until further, more comprehensive, sampling is carried out in Australasia this remains a murky area.
There is no meaningful evidence that all mtDNA and Y chromosomal data roots in Africa not Australia. Your understanding is based on a glaring assumption, that haplogroups L3, M, N and CT emerged first in Africa around 73,000 years ago. There is absolutely no proof of this, the oldest African DNA samples are a mere 8000 years old. The consensus view is based on sampling African people from the present day and then assuming all the direct ancestors of these people (myself included) lived in Africa from the present day all the way back to hundreds of thousands of years ago. This is a huge and wild assumption. The evidence does not fit well with this at all, hence recent papers that have argued for HgL3 emerging in Asia and moving into Africa - something I already had explained in my book. The mutation rates of Y-chromosomal material and mtDNA differ by 10X, the likelihood of mutations on both lines occurring at the same moment (73,000 years ago) in teh same place (East Africa) are incredibly small. The better fit has always been a migration across the Bab-el-Mandab carrying these haplogroups into the continent. Today we now know that there was a climatic disaster underway outside of Sub-Saharan Africa with every reason for humans in western Eurasia to head towards the safe-zones of sub-equtorial Africa.
Did you know that the oldest variants of HgM and HgN are all found among Aboriginals?
“The southern route hypothesis proposes that the Eurasian branches (M and N) of the macrohaplogroup L3 differentiated in or near the African continent and rapidly spread across the Asian peninsulas to reach Australia and Melanesia. Under this assumption, it is expected that, in general, coalescence ages of haplogroups should decrease from Africa to Australia. However, we have demonstrated that this is not the case. Just on the contrary, the oldest M and N haplogroups are detected in southern China and Australasia instead of India, and associations between longitudinal geographic distances and relative ages of M and N haplogroups run, against to expectation, westwards with younger haplogroup ages going to Africa.” [www.biorxiv.org]
It should also be noted that while HgL3 and both HgM & HgN are closely related, they are considered to have emerged at essentially the same time, it has never been explained how that can work, why assume HgL3 is ancestral to M & N? The answer is it is very convenient for Recent Out of Africa but makes no sense in terms of the mutation rate of mtDNA. These three lines did not all magically appear suddenly 73,000 years ago, they were driven west by the climate and therefore appear in the genome of early Africans close to that time.
It is simply a case of a 'garbage in, garbage out' model. Yes, if we place the origin of all these haplogroups in Africa without any compelling evidence, and a great deal of contrary data, then we can make all the rest of the model sound very solid. The entire thing is built on quicksand.
There is more and more doubt that AMH emerged from Africa with archaic Homo sapiens in 260Kya and fully modern humans there at least 120Kya. These early migrants have left very little genetic legacy because of the climatic event that reduced them to a handful of survivors and pushed many of them into Africa.
The hypothesis that these founder haplogrups are African is baseless, wrong, and in collapse.