I found a study in my saved files that looked at the global distribution of commonly tested snps for skin de-pigmentation.
The full Norton et al paper can be found here:
http://mbe.oxfordjournals.org/content/24/3/710.full
This handy table lists the snps, the gene in which they appear, and the global distribution. The MATP gene is the one in which SLC42A5 can be found. SLC24A5, SLC42A5, and TYR together account for the vast majority of the variation between SSAs and Eruopeans.
http://mbe.oxfordjournals.org/content/24/3/710/T1.expansion.html
This table provides the exact percentages in list form for five snps by Hap Map population.
http://mbe.oxfordjournals.org/content/suppl/2006/12/21/msl203.DC1/mbe-06-0529-File010_msl203.
I checked for all of them in the supplement of the Olade paper, and other than ASIP rs6058017, which I couldnt find, the others were tested, and La Brana was ancestral for all of them.
I also found it helpful to read the following forensics paper, where they make it clear that in order for the probabilities to be reliable for lighter skin pigmentation, the sample must be homozygous for three certain high value snps. Even for medium skin, the sample must be homozygous for at least two of a set of snps. When all criteria are met, the false prediction rate is 1%. The forensics test uses some of the snps used by Norton et al, namely SLC45A2 and 24A5, but not all of them, and includes other snps associated with pigmentation.
https://ncjrs.gov/pdffiles1/nij/grants/242774.pdf
From the forensics paper... Non-dark skin color (i.e. light or medium) is predicted by any two of the following
alleles, GG at rs12913832'Herc2, GG at rs16891982'SLC45A2, AA at rs1426654'SLC24A5, TT at rs1545397'OCA2, or AA at MCIR rs885479.
Light skin color is predicted by more stringent conditions, GG at rs1291382, which is the Herc2 gene, plus GG at rs16891982 which is SLC45A2, and AA at rs1426654, which is SLC24A5. All three must be present and homozygous.
Non light skin color, i.e. medium or dark is predicted by GG at rs6119471 on the ASIP gene.
As for La Brana, he is derived homozygous C on rs6119471 on the ASIP gene, which is considered a weak effect gene. He is also homozygous for the derived GG on the Herc2 rs12913832 snp, and heterozygous T on IRF4, rs2203592, which is considered a medium effect gene. So, he has a total of 5 out of 14 possible snps on the test, although as used, there is a qualitative criteria instead of a merely additive one.
As the authors mentioned in the text of the paper itself, La Brana is also heterozygous derived C for TYRP1, rs1408799, which is also implicated in relation to eye color.
Applying the standard forensics criteria, he would be predicted as having dark skin.