The major premise of the Fourth Way is the only one that raises hackles:
More and less get said about diverse things insofar as they approach something maximal.
Magis et minus dicuntur de diversis secundum quod appropinquant diversimode ad aliquid quod maxime est.
One argument for the premise comes from what we might call the direction of recognition. Take a series of things that are more or less perfect, like three triangles more or less perfectly drawn. You recognize a better and worse, but in what order? It seems evident that you take an ideal and measure the degree to which each falls away from it rather than taking a bad triangle (or, what’s much less likely, the perfectly awful triangle) and remove its badness from it. You can touch up a picture by taking away its imperfections, but all our argument requires is that such touch ups are attempts to approach an ideal, not attempts to gain distance from an anti-ideal. The reason is that ideals are definite in a way that allows for an approach whereas failures to instatiate something don’t form a definite point of reference that we could distance ourselves from. A sloppily drawn face, a straight line, the spelling of my middle name, and a perfectly drawn hand can all be failures at portraiture, but these don’t coalesce into some definite point of reference from which we could remove imperfections. Logically, any recognition of an ideal from its inferiors happens only after some recognition of the ideal itself.
Judging something from an ideal does not require that we have a perfectly clear view of it – in the actual progress of discovery we usually tinker around to see what works rather than executing a perfectly top-down instatiation of a perfectly grasped ideal. But this tinkering and discovery doesn’t change the logical priority of ideal instantiations.
If this is right, then Descartes is vindicated in his claim that the infinite good is known before the finite, even if it is not the first thing distinctly known in the order of discovery; and Aristotle’s theory of abstraction is a way of fleshing out what something is whose existence is already known in another way.