L3 -> M: Ancient linkage of african and south asian populations?

My DNA results came, my mtDNA is some unspecified L3 variant. As it was a low-hes test, I'm even thinking it could be a M mtDNA clade, due M would be a direct derivative of L3 lineage.

M mtDNA is a well rooted maternal haplogroup in South Asia, with an important prevalent in East Asia and Oceania, but it also occurs in some extant in Africa, almost always moving together with L3 clades, what makes we think sometimes L3 also is a result of a back-to-Africa migration and we even can push many L3 clades to M mtDNA.

Had also someone this inference?

https://upload.wikimedia.org/wikiped...condriales.PNG

One should think on the reason why the mtDNA of the ancient Leang Panninge from Sulawesi from the vicinity of the Sunda shelf was determined to be mtDNA M1 by one piece of software, but determined to be M* sitting on one branch with a Papuan mtDNA Q representative by another piece of software.
Interestingly, it is mtDNA R23 that is a Sunda shelf-related lineage without apparent Denisovan and Neanderthal mutations in its basal sequence. The mtDNA R23 was also observed in a Cambodian (Khmer) (Cambodia_HGDP00719 Cambodia Cambodian Cambodia Cambodia HGDP00719 R23 Cann et al., 2002).

Waves of distribution of mtDNA M and mtDNA R bearers in Southeast Asia/Papunesia

Genome of a middle Holocene huntergatherer from Wallacea
https://media.springernature.com/ful..._Fig1_HTML.png
Mitochondrial analysis
All unique mitochondrial reads (16.72x coverage) were used to reconstruct the mtDNA consensus sequence and estimate mitochondrial estimation using schmutzi70 with quality filters 0, 10, 20 and 30. Mitochondrial contamination was estimated to be 2 ± 1%. Since the mitochondrial/nuclear ratio of the shotgun sequence was low (61.78), the mitochondrial contamination can serve as a proxy for nuclear DNA contamination . The mitochondrial haplogroup was ascertained as M1 using Haplofind72 but with a low confidence score (SI Table 17).

The q30 consensus sequence was aligned against the consensus sequences of 53 present-day human and several ancient mitochondrial genomes from China73 , Siberia74-76 , Mongolia77 and Southeast Asia78 using MAFFT79 . We then constructed a maximum parsimony tree, eliminating all positions with less than 97% coverage (~2 bp missing/position) and calculating 500 bootstrap replicates in MEGA v.10.1.5 (ref. 80). The Leang Panninge mitochondrial genome falls basal of a haplogroup M clade, but on a different branch than the Hòabìnhians78 (SI Fig. 2).

5.10 Haplogroup R23
Figure 5.45 shows R23 is a rare haplogroup and undergone high drift resulting in a date of ~9 ka. It is represented here by two complete mtDNA sequences from Vietnam (Peng et al. , 2010) and Sumba of Lesser Sunda Islands, Indonesia (Archaeogenetics Research Group, Huddersfield). The HVS-I data showed that R23 is seen in Sumba and Bali, Indonesia (Hill et al. , 2007), suggesting that this rare relict subclade most likely has a root on the Sunda shelf.
https://etheses.whiterose.ac.uk/7872/