Godfrey I would suspect that if there is a logical inconsistency here, it would be in DeWitt's description of "perfection" rather than in Epicurus' theory. Surely you are right that in fighting Plato's ideals, Epicurus would not then turn around and adopt the same problem in a new form.
No doubt there's a lot more investigation to be done into the background of the isonomia discussion, but I would expect this reasoning to be of the "nature never makes only a single thing of a kind" variety that is clearly described by Lucretius.
In that other case, it seems they were taking from the observation that here on earth there is nothing absolutely unique, but that everything belongs to a class of similars, and extrapolating that to the universe as a whole. That's a reasonable procedure it seems to me, and the burden of proof to refute a theory based on what we see here on earth is on someone who asserts that there IS something that is absolutely unique but which cannot be observed (a burden which cannot be met by observation) rather than on Epicurus' position, which rests deductions from what we do in fact observe.
We'd have to dive into whether in fact Epicurus was talking about "perfection" as a concept, but I would suspect more that he was (1) starting with the observation that there is nothing unique, all belonging to a class of similars, and then proceeding to (2) that within any class of similars, it is possible for us (probably by process related to preconception) to "rank" those similars in terms of varying characteristics.
Happiness and length of life being characteristics of the class of "living beings," it makes logical sense to conclude from the examples we observe that happiness and length of life exist on a spectrum from less happy to completely happy, and from short life to much longer life. Given that we see such spectrums of characteristics here on earth, it is logical to believe that in a universe that is eternally old and boundlessly wide, there exist examples within those classes that extend out in both directions (greater toward total happiness; longer life toward deathlessness).
We could probably compare what Philodemus is saying in "On Signs," but it does seem that the Epicureans were taking the canonical position that it is legitimate to reason in that way, from things that we observe here on earth to how those things might occur in wider variety in an infinite number of places.
Remember too, that the purpose of scientific investigation is not to latch onto a single true theory of everything, which is essentially impossible, but to determine what is both consistent with what we observe and not contradicted by that which we observe. Then, using sound reasoning, we then consider what is possible to be a real possibility, while "waiting" to reduce the number of possibilities if any new evidence arrives to contradict one of more of them.
This is very different, and in my view superior, to the modern arbitrary speculations that sometimes seems to be "if I can't see if for myself here then it didn't and doesn't happen," or that accept some theory that has absolutely no evidential proof behind it, but which makes sense in some logic-only analysis.
So I see this as an area where DeWitt has generally pointed in the right direction, but further research needs to follow his path and extend his observations more precisely.
And yes it's definitely a process that has limitations, but I see this as very similar to logically dividing all experiences of life into pleasure and pain, and then logically observing that the presence of one is the absence of the other. Epicurus wasn't willing to use logical processes to work with Platonic ideals that have no evidence in reality to support them, but he was willing to use logical processes to work with the evidence that the senses provide to us, and Epicurus seems to have been just as sharp a logician as any of his enemies, carrying his reasoning to its logical conclusions.