The literature on the seventy-two is very large and very uneven. A few habits make it navigable.
Date every claim. Ask which century a given detail comes from. Weyer is 1577. Scot's English is 1584 and the edition that matters is 1665. The Lemegeton's manuscripts run from the later seventeenth century into the nineteenth. Rudd's angel pairing is seventeenth-century. The decan, tarot and elemental correspondences are Golden Dawn, from the 1880s. The familiar pictures are Le Breton's, from 1863. These layers are routinely printed as one.
Go to the manuscripts where you can. Sloane 3825, Sloane 3648, Harley 6483 and their relatives survive in the British Library and have been edited and published. A striking proportion of confident statements in secondary sources cannot be traced back to them.
Distinguish contested from unsourced. Whether Astaroth derives from Astarte is genuinely contested and should be reported that way. Whether the Lemegeton had one compiler is genuinely open. But a claim that a spirit governs a particular hour, or answers to a particular incense, or has a documented traditional attribute, is usually not contested at all, it is simply unsourced, traceable to one twentieth-century book that has since been copied without attribution.
Treat confident specificity as a warning sign. There is a commercial market in this material, and detail sells better than uncertainty. The primary texts are far barer than the secondary ones.
Do not mistake coherence for provenance. The seventy-two make an elegant system once the angels, the decans and the tarot are laid over them. That elegance was manufactured, in stages, by people whose names are known.