Earlier quoted context omitted.
It's important to note that [1] seems to refer to federal sentencing, and almost all criminal law in the U.S. is state law. [1] is a small subset of what's going on, and it can't be used to generalize.
Not the OP, but it is interesting that the US criminal law is so state dependent. Out of curiosity: do you have reason to believe that the sentencing disparity is better when considering states?
Statistics is hard, and the entire reason there's leeway in sentencing is to account for situations which are resistant to statistical regularity.
There is, or was, a pretty obvious source of disparity in US law: freebase cocaine was a felony at much smaller quantities than cocaine salt, and the former was more likely to be used and dealt by black Americans.
Would that show up in such an analysis? They were different convictions, after all. So the only answer can be "yes or no", and so on ad infinitum.