Updated and bumped: Feb-1-09. New paragraphs in ochre red.
Let's try to approach the issue of the human Eurasian expansions (adressed in previous posts via analysis of mtDNA structure) visually. Lines are used instead of dots when the geographical scope of a clade is large and the core of it can be argued.
1. 1st expansive pulse (maybe soon after Toba super-eruption, c. 74,000 BP):
This map shows the clades that (at their diversification moment) are separated from the L3 node by 5 or 6 SNPs (i.e. right after the formation of M and N). Clades in red are descendant of M, while those in blue are descendant of N. M and N as such are not shown but can be infered from their derivates.
If we could divide this pulse in fine detail following strictly the number of SNPs (something I am not sure it's realistic), we would get:
1.1. M node (preliminary M expansion/diversification).
1.2. N node. Simultaneously expansion of 5 M subclades, notably in SE Asia (M21 and M9-E). Other haplogroups expand (avant-guards?) in each of the three neighbouring regions. 1.3. Expansion mostly in South Asia (4+ subclades) with offshots also in all neighbouring regions: East Asia (M7), Australia (S) and West Eurasia (macro-X-N1).
The whole pattern argues for a homeland either in South or SE Asia. Or maybe at both regions at the same time.
There is no clear difference between M and N patterns, except that, at this stage, M is clearly present in SE, East Asia and Melanesia, while N is not (but is in Australia and West Asia). This could suggest a more westward tendency for N since the beginning and maybe an association with A blood type (most common in West Eurasia and among Australian Aborigines precisely).
In any case both macro-haplogroups moved through South and SE Asia. There is no clear barrier.
2. Interlude: brief pause of expansion and formation of haplogroup R:
This map shows the clades separated from L3 by 7 SNPs. They seem to form a narrow valley in the process of expansion/diversification (see previous post) and the moment is also worth a separate snap because it is when macro-haplogroup R was formed.
This is the first moment we see any N-derived presence in East Asia. This fact would argue for a South Asian (or at most SE Asian) origin for the R-N9 super-clade.
3. 2nd expansive pulse (some 60-50,000 years ago?):
This map shows the main clades that are separated from the L3 node by 8-10 SNPs. There is a peak of haplogroup nodes at 8-9 SNPs, which I understand as the second expansive pulse or early Eurasians. The 10 SNPs mark would rather be the end of that pulse and, in general, of the main expansive period.
In this and the next map, R derivated clades are marked in green, while blue is reserved for the N(xR) haplogroups.
We can again take a risk and divide this epysode in three possible "moments":
3.1. The first and main expansion of R (5 clades) reaches from India (R30 and R31, the latter reaching also to Australia) to New Guinea (P - seems I forgot to draw it in the map, sorry), including quite naturally SE Asia as well (B and macro-F). Simultaneously, in the M family, there is expansion in East Asia (M8-CZ) and of a rare clade (48) found at least in Saudi Arabia.
3.2. This subphase (the most expansive "instant" of the whole process) would rather belong to M, with four clades expanding in South Asia (M2, M35, M39 and M40), one into Andaman islands (M31), one in East Asia (G) and yet another one in all Sahul (Q). The expansion of R0-HV into West Asia also appears to happen in this moment. In East/NE Asia we also see the expansion of CZ (separated already from M8)
3.3. Less important "photogram". It includes three apparent haplogroup expansions: M34 in South Asia, M42 in Australia and U (R-derived) between South and West Asia. It seems to be fully within the line of the previous "moment" and includes the last significative expansion into Sahul before Modern Age.
Overall the phase appears to show a fast expansion of R from India to Sahul, followed by an even more impressive second expansion of M in all the regions except SE Asia. Also R-derived clades begin making clear inroads into West Asia.
4. Aftermath, more diffuse expansion through a longer period:
This map shows the main events between 11 and 15 SNPs after the L3 node, including most of the expansion into West Eurasia and North Asia. Events after that I'm ignoring here.
The most eye-catching features here are that Eastern Eurasia and Sahul appear now disconnected from South Asia, which still shows some (weak) connections with West Eurasia. We can consider this phase as the beginning of "racialization" therefore, once the different Eurasian regions appear to have become mostly autonomous from each other.
It is also noticeable the inroads into Europe and North Africa by West Asian lineages at this rather extense phase.