Apportionment of Human Genomic Variation: from Lewontin’s fallacy to Edward’s paradox