The questions actually aren't much difficult to answer. Nilotic Kenyans are lightly built (with low BMI), with long, slender legs that contribute to their excellent running economy. Ethiopians and Somalis (Cushites) are even more slender, which suggests that their performance in the 3000 m steeple and the 10 000 m (where BMI is the lowest) should be even better than in Kenyans (which is largely true in the latter case, but not in the former, apparently due to the lack of tradition). In the marathon, the trend towards lower BMI values again reverses, because the ability to store energy becomes equally important like running economy (more muscle=more glycogen=higher speed throughout most of the race). Hence marathon runners are heavier than 5-10 km runners and they actually resemble stunted milers. This can also explain, why non-Africans are more competetive in the marathon than in the 5-10 km.
The body types of Nilotes and Cushites don't fit so well in the 800/1500 m, however. First, these distances require a much higher body height than we encounter in the famous "running tribes", and second, the height/BMI relationship in these distances is the exact opposite of what we see in East Africans: taller 800/1500 m runners tend to be heavier, because they rely on speed and anaerobic capacity. In contrast, BMI in East Africans progressively decreases with body height (Bergmann's rule), which means that at the body height necessary for the 800/1500 m (ca. 180 cm), they would be too thin and they would probably lack most anaerobic prerequisites for these distances. This can also explain, why so many East Africans tested positively on steroid use. (It may seem bizarre that steroids would help distance runners, but in this case, it could apparently be beneficial).
Historically, Kenyans have been more successful in the 800 m than in the 1500 m, which is a physiological nonsense. This fact indicates that the 1500 m has always had a much higher prestige and hence the level of international competition was much higher. In contrast, the 800 m has been a "poor neighbour" of the mile.
The West African sprinting dominance is equally easy to understand. Laboratory testing demonstrated that West African students in Canada had 9.3% higher proportion of fast-twitch fibers in the vastus lateralis of thigh muscle than white Canadians (68.3% vs 59.0%). Similarly, black university students in USA had 7.9% more fast-twitch fibers (62,8% vs. 54,9%) than white students. Besides that, some anatomical characteristics like short, massive thighs and long, thin calves with lower inertia during fast movement can contribute, too.
What is bizarre, however, is that people read these numbers and deduce that they guarantee blacks an equally strong advantage in all distances that we call "sprints". This is wrong. First, the 400 m distance has fundamentally different physical and physiological determinants of performance than the 100-200 m. The runners here are much taller (about +5 cm) because their performance depends on anaerobic capacity that increases with body size (=stride length). Naturally, the height factor alone favours Europeans, because they are by far the tallest humans on Earth. Further, although this distance is still overwhelmingly anaerobic, the proportion of aerobic energy is almost 3-times higher than in the 100-200 m.
It is unfortunate that I couldn't find any trustworthy data on the typical fast-twitch muscle fiber composition in 400 m runners, but at least, I estimated it from their typical VO2 max. values (around 60 ml/kg.min), based on equations developed from studies of endurance runners. I got 58% fast-twitch fibers. For a layman, it would be a shocking finding, because it would mean that at worst, white 400 m runners shouldn't have worse physiological prerequisites for the 400 m than West African blacks. And while in men, we still could discuss some nuances of these numbers, in women, the situation is rather clear: The average proportion of FT fibers in Europeans (55% or slightly more) corresponds to a long sprint lasting something below 1 minute.
Second, the height advantage of Europeans should translate even into the 110 m hurdles. In fact, the average height of international medal winners from the period 2000-2013 is 186,7 cm. In the 100 m, it was only 181,2 cm. Apparently, a sprint is not always the same sprint.
Therefore, based on these data, it is quite clear that we shouldn't make much fuss around the current situation in the 400-1500 m. It is hilarious and it has nothing to do with the real distribution of global athletic talent. Consider, for example, that the degree of dominance of West Africans in the 100 m and 400 m is almost the same (which is utterly absurd), and out of mere two white 400 m runners among historical Top 30, one set his personal best in 1976. If you take these stats seriously, you are a fool.