The following two claims seem plausible enough to me:
1. God is not morally obligated to create the best possible world.
2. There are no supererogatory acts.
Supererogatory acts are those acts that go above and beyond what duty or obligation requires. But if God isn't obligated to create the best possible world, and is merely obligated to produce a good enough world, then isn't it better if God creates a world that's better than the minimally good enough world? It seems like a supererogatory act for God to create at all, since it will never be the best act of creation. So there does seem to be a problem if you accept both these claims. But, though I would not submit to martydrom for either claim, there do seem to me to be good arguments for both, and yet they seem inconsistent.
1. I think it's plausible that adding one more intrinsically good thing to a world will make the world better, and its always possible to add one more intrinsically good thing. This means there is no best possible world, and thus it is impossible even for an omnipotent being to create the best possible world. Unless God is obligated to do the impossible, it seems that claim 1 is true.
2. Consequence-based ethical theories have usually required maximizing the best consequences, but a lot of people have rejected such an approach, because it implies that it's wrong to go see a movie because that money could better be spent helping starving people get some food (for one example). So we now have satisficing theories approaches that say that all we're obligated to do is seek good enough consequences. A similar approach occurs in non-consequentialist ethics, where perfect duties are duties everyone has but imperfect duties are acts that someone or other ought to do but no one particular person is required to do them.
We usually take supererogatory acts to be those acts that go above and beyond what duty or obligation requires. Someone can meet all duties or obligations but still be able to do more good than is required. Such acts would be morally better than the acts duty or obligation requires, and thus a person who does them would be morally better than a person merely meeting all obligations or duties.
I don't have a good philosophical argument for why there are no supererogatory acts for humans, but I do think it follows from Jesus' teachings. He taught that we ought to go the extra mile, turn the other cheek, give the shirt and not just the asked-for cloak, etc. It's not just a recommendation to do more than seems morally required. It actually is morally required. So Christians at least have good reason not to believe in supererogatory acts for us.
That's not a philosophical argument. But it's always struck me that the idea of supererogation is often just an excuse not to be good enough, sometimes even to avoid clear moral obligations. For example, Judith Jarvis Thomson uses it to argue that it would be perfectly fine to kill your own offspring at a stage when that offspring has full moral status and is dependent on your body, as long as you made some reasonable attempt to prevent that person's existence but knew your freely-chosen actions could nevertheless result in such a situation; Thomson's principle actually implies the conclusion that you have no obligation to care for a baby left on your doorstep or even to inform anyone about it so they can do so. But you can probably accept some supererogation without the monstrous conclusions that follow from the principle Thomson uses to explain her acceptance of supererogation. So I don't think this kind of consideration will necessarily support the claim that there is never any supererogation.
Nevertheless, I do have a philosophical argument for 2 if we restrict ourselves just to God. A perfect being is perfect by nature. God will only do what's consistent with his nature. God won't be more perfect by creating a world that's a little better. So it doesn't seem as if supererogation applies to God. There are no actions that are better to God for do, with other actions merely being less good but morally allowable.
It occurs to me that this way of removing supererogation actually doesn't lead to the inconsistency, though. One way to remove supererogation says that we ought to do the best possible. But this way of removing it says not that we ought to do the best actions possible but that we ought to be the best possible person we can be and do actions consistent with that best moral character. A character-based approach to ethics (as opposed to an act-based approach) will thus think of supererogation differently enough from how we typically do, given the overwhelming influence of act-based ethics, and I think it actually removes the original inconsistency I was proposing above.
A character-based approach to supererogation says we ought to have the best character possible, which on the human level explains why doing lots of good is never enough, and I think that can ground the kinds of ethical claims Jesus taught. But it's not the sort of view that requires maximizing good consequences, and it seems to me to be perfectly compatible with thinking that there is no maximum good world. Supererogation may seem like an excuse not to do what's best, but if the issue is being the best person in terms of your character, then you will seek to be best without its being grounded in doing the best actions. The influence is the other way around. If you are good, then you will do good things because you are good. A perfect being will always act with perfect wisdom and goodness and can be said to act perfectly, even if there is no best outcome out of all the possible outcomes God could consider actualizing. So I think you get satisficing with respect to the best possible world. There is no best possible world for God to actualize. And yet it's not because God only has to be good enough. God will be perfectly good either way. That perfect goodness can result in any of various possible levels of good in the world. The consequences of God's acts aren't what make God good. Rather, a good being will do good if that being creates at all, but God would still be good if he didn't create at all.
Of course, if you take God's perfect nature to be infinitely good, then it doesn't matter how good or bad the finite goods of the created universe are on a consequence-based ethical view, because the universe isn't any better with more good in the world and isn't any worse with less good. So if I became convinced that my proposed solution to the inconsistency won't work out, one way out of the problem might be to say that this is a maximally-good world if you include God's infinitely-good nature in the calculation, and thus even if God created a world that, taken in itself, isn't as good as another, it's still true that the entire situation (created world + God) is infinitely good in a way that can't be greater or less than any other situation (given that God's existence is necessary).
So I think I can actually maintain both claims without any inconsistency arising, at any rate.