For Ust'-Ishim (the oldest modern human human genome yet available and the oldest conclusively modern human remains from northern Eurasia), Fu et al. 2014 report a 95.4% radiocarbon confidence interval of 46,880–43,210 calBP and a Neanderthal admixture date estimated at 232–430 generations (6960-12900 years at 30 years/generation) upstream, hence a lower floor of 60 kaBP for Neanderthal admixture into this lineage – in the same stroke, non-Africans broadly. A human presence in Australia 65 kaBP or even 60 kaBP immediately suggests discontinuity between the initial settlers of Sahul and its present-day inhabitants, who bear the impress not just of Neanderthal admixture but also a later Denisovan input. There are a number of ways to suture the discrepancy shut, but I will walk through some alternatives conditioned on 50–60 kaBP being the correct range for Ust'-Ishim and extant Eurasian lineages.

Sankararaman et al. 2016: “To obtain a rough estimate of the date of Denisovan admixture—cognizant of the fact that for Oceanians we do not have the information needed to fully correct for uncertainty in the genetic map—we calibrated to previous estimates of the date of Neanderthal admixture, under the simplifying assumption that the date of Neanderthal admixture in the history of New Guineans is the same as the 50,000–60,000 years ago estimated for a radiocarbon-dated Upper Paleolithic Siberian [3]. Rescaling by 1,000/1,121, we estimate 44,000–54,000 years ago for Denisovan admixture.”

If we accept the new Madjedbebe dates but posit that the people responsible for the oldest occupation were ancestral to present-day Aboriginal Australians, we're forced to reject this simplifying assumption. If we for instance place Denisovan admixture immediately prior to a 65 kaBP arrival in Australia, we "backdate" their Neanderthal admixture to roughly 72.9 kaBP (66.5 kaBP with Clarkson et al. 2017’s “conservative” Madjedbebe estimate of 59.3 kaBP; whenever it happened, it is obviously younger than the Neanderthal input, with the estimated generational scale factor noted above). Retaining the younger date estimate for Neanderthal admixture in Eurasians introduces a serious Oceanian/non-Oceanian discordance in histories of archaic admixture that is implausible for a number of reasons. Importantly, it would be difficult to reconcile such a scenario with how much genetic history present-day Australians and Papuans clearly share with East Eurasians, who most parsimoniously received their Neanderthal admixture in the same main event – presumably somewhere in the Near East upstream of the clear differentiation of these extra-African lineages – as Ust'-Ishim and ancestral West Eurasians. (The recurrent proposals of deeply divergent "Southern Route" ancestry in Oceanians were severely checked by Mallick et al. 2016.)

Alternatively, with fewer contortions, we can picture a significant genetic discontinuity between the initial inhabitants of Australia and present-day Aboriginal populations (no contribution, or something verging on it in subtlety). Should that be the case, we lose our ability to constrain the Madjedbebe people's position in the human graph.