Probabilistic and parsimony-based arguments regarding available genetics data are used to propose that Hardy and Higgin's amyloid cascade hypothesis is valid but is commonly interpreted too narrowly to support, incorrectly, the primacy of the amyloid-β peptide (Aβ) in driving Alzheimer's disease pathogenesis. Instead, increased activity of the βCTF (C99) fragment of AβPP is the critical pathogenic determinant altered by mutations in the APP gene. This model is consistent with the regulation of APP mRNA translation via its 5' iron responsive element. Similar arguments support that the pathological effects of familial Alzheimer's disease mutations in the genes PSEN1 and PSEN2 are not exerted directly via changes in AβPP cleavage to produce different ratios of Aβ length. Rather, these mutations likely act through effects on presenilin holoprotein conformation and function, and possibly the formation and stability of multimers of presenilin holoprotein and/or of the γ-secretase complex. All fAD mutations in APP, PSEN1, and PSEN2 likely find unity of pathological mechanism in their actions on endolysosomal acidification and mitochondrial function, with detrimental effects on iron homeostasis and promotion of "pseudo-hypoxia" being of central importance. Aβ production is enhanced and distorted by oxidative stress and accumulates due to decreased lysosomal function. It may act as a disease-associated molecular pattern enhancing oxidative stress-driven neuroinflammation during the cognitive phase of the disease.