Source:
The human genome is characterised by many runs of homozygous genotypes, where identical haplotypes were inherited from each parent. The length of each run is determined partly by the number of generations since the common ancestor: offspring of cousin marriages have long runs of homozygosity...
journals.plos.org
Most of the world has ROH Mb lenght of 50+, making Somalis with their 52 ROH Mb totally unremarkable and not inbred by global standards.
It's just that other Sub-Saharan Africans are odd with their very low ROH mb length.
You are so wrong, I don't know where to begin. The 52.2 number is a count of how many runs of homozygosity (ROH) exceed 1.5 Mbs. Your image is a count of ROH between 0.5-1 Mbs. It is totally irrelevant to this discussion. The number of ROH in this range is useful for detecting founder effects, but it is not very useful for detecting recent inbreeding.
Somalis, Oromo and Amhara all clearly exhibit a founder effect in comparison to other Africans. This is likely due to the fact that all three have significant Eurasian ancestry and exhibit a portion of the OOA founder effect present in all non Africans. You can see this most clearly when you compare Horn Africans to other Africans.
Notice how high the Amhara, Somali and Oromo are in mean total ROH 0.3-1 Mbs? This is because of the Eurasian bottleneck. But because the Amhara and Oromo practice exogamy, their ROH above 5 Mb in length are many times fewer than Somalis. ROH greater than 5 Mb are almost always a sign of inbreeding. You can then see the spike in ROH greater than 8 Mbs among Somalis, i.e very recent inbreeding.
Take a look at the first column ROH less than 1.5 Mb. Somalis, Oromo and Amhara have values much higher than other Africans because of Eurasian ancestry. Somalis at 146 and the Kikuyu at 85 for example. Low effective population size does not explain why the mean FROH of Somalis is 7x that of the Amhara, your explanation does not even come close. The Amhara and Oromo do have larger effective populations than Somalis but just barely, and certainly not enough to explain why 48.7% of Somalis have a FROH that is greater than second cousins.