It's an insincere argument
It's not reasonable to connect vitamin D to anything 100k years ago when vitamin D is so heavily influenced by diet and weather, both of which changed wildly. Lactose tolerance, megafauna, fishing, the last glacial maximum- just 20,000 years ago people were undergoing massive differentiation in vitamin D production and utilization.
Shorts and t-shirts? Em, a rather large amount of modern urban African spends most of their day indoors and wears long-sleeved shirt/pants/long dresses/long robe/hijabs/etc.
I'm not sure why one would expect them to have >50% of prehistoric levels.
The paper you linked to (or really, the metaanalysis it cited) showed low D levels were associated with urbanization, but the lowest concentrations observed in northern African countries and in South Africa with seasons. The highest levels were among those practicing traditional lifestyles.