You should also read, "How Can We Be Moved by the Fate of Anna Karenina?" by Colin Radford, and the literature that developed in response to it.
I think a lot hinges in your question on the word "slightly". Is there a moral obligation to keep a promise to do something that is "slightly immoral"? I think that the answer has to be "No", since the value of duty to keep promises is not in question, and the act contemplated is only "slightly" immoral. OK, but how slightly? Would it help if you had written, "if I promise to do something that is utterly and completely immoral"? Or if you had written, "If I promise to do something that is only ever so slightly, just the teeniest barely discernible bit, immoral"? I think such gradations make a big difference, and it is not very clear how "slight" the immorality has to be before it ceases to conflict with the important general obligation to keep promises. Of course much depends also on the question to whom the promise was give, why, under what circumstances, and so on. These all need spelling out before we can address the question with any hope of answering it.
Consider two otherwise identical worlds, in one of which determinism is true, and in the other of which it is not. In the second world there is freedom of the will. What difference does this make to the legal system and the moral system? Ex hypothesi, none. When I wrote that if hard determinism is true, moral systems cannot be corrupt, I was describing Stephen's view, not my own! Hence I think the apparent contradiction that he detects. Here's another way of putting the point I wanted to make. We can imagine a deterministic world in which people are thinking good thoughts and doing good things, and making good laws. The world of law and morality seems to have survived the loss of freedom! In this world, some philosophers believe in determinism, and some in freewill. But of course the having of those beliefs is part of a deterministically produced system. There could be incidentally be a deterministic world that is morally and legally better than a world with freewill. Consider the worst world imaginable, but having freewill, and the best world imaginable, but lacking freewill. You might prefer to live in the unpleasant world, but that doesn't mean it's better. In it the innocent are tortured, unfairness abounds, and so on.
I think the answer may be that phenomenology has produced so disappointingly little. In a non-philosophical sense phenomenology is defined as the preliminary classification of phenomena in an enquiry. So one might for example regard it as a piece of phenomenology in this non-philosophical sense to say that a white surface seen through a light blue filter looks stone-cold white, and not blue at all, as per the philosophical folklore. The question the analytic philosophers ask themselves, I suspect, or at least this one does, is whether there is something as solid and productive that can be gleaned from phenomenology in the philosophical sense, in addition to its methodological meanderings.
I don't think much depends here on whether psychology or psychiatry is a science. What matters is whether they are doing good or not for patients. Supposing we think they are, a commonsense criterion seems to be that even if patients do not accept the diagnosis or the treatment, things change if they are contemplating harm to themselves or others, as the jargon has it. The jargon persists and is more than jargon because it has a real role; it is a criterion. A second criterion or perhaps test might be this. Can the patients get through the day or week to their own satisfaction, or look after themselves OK? If I have a friend who never gets up, doesn't bathe, won't look me in the eye, can't work and doesn't eat, and has gone down to eighty pounds, then I feel justified in intervening, because of this second criterion, over the stated wishes of the patient. Something is wrong, and I have an obligation, or some responsibility anyway, to help, especially if I am a psychiatrist. Most people in trouble do want help, I think, even if we are too dim to find the way. "Normative" ethics means ethics which is not-metaethical. I think you may be misusing the word; what you may mean is a highly directive ethics that conflicts with the patient's own ethics. The patient too has a normative ethics - his own. And what if psychiatry is misguided or "pseudo-medical" and does harm? If that is true, then we should avoid it, of course. But is it true? It seems to me it is a bit improbable that the profession as a whole is phony. It is not monolithic enough for that, for one thing. One can always switch psychiatrists.
The principle of sufficient reason, due to Leibniz, states that there is always a reason why some particular thing happens, rather than some other thing. This does not immediately or obviously pose a threat to freedom. Note that "reason" does not mean the same as "cause", although a cause might be a reason.
Determinism states something much stronger, more complicated, and more sinister. It tells us that the laws of nature and the initial state of the universe at some time in the past entail the state of the universe in the present. Entailment is a strong relation, and what determinism means is that if the laws are whatever they are and the initial state of the universe is whatever it is, then the the universe must (nota been, "must") go into the subsequent state. There is a necessary truth. It is that if the universe is in the initial state, and the laws apply, then the universe will go into the second state. Determinism has been held to pose a severe threat to freedom in the metaphysical sense, or freewill, though so-called compatibilists have a view which de-fangs determinism, if it is successful.
The principle of universal causation is different again. It states that every event has a cause. This might well be false of the first event in time, if it really is the first event, because there is no prior event available to cause it, since nothing is prior to the first event.
Stephen is right. We should distinguish the Design Argument from the Ontological Argument. Your question concerns neither. Your question is about the Problem of Evil, so called. How can a being who is all-good, all-powerful and all-knowing allow evil to exist? The simplest way to solve this problem is to deny one of these three propositions, and it is perfectly acceptable to deny the second: God's power is limited. This approach is taken by process theologians, who say that God is developing. For the typical process theologian, as for the Mormon, God cannot break the laws of nature, for example. The trouble with this solution is not that it does not work for the theist. The problem is that it does not work for the traditional Christian theist, and as far as I know also for the Jewish and Muslim theist. A god with limited powers is simply not recognizable as the Creator of the Universe, the Father Almighty, and so on. So the solution is logically acceptable, but theologically unacceptable.
There are numerous complex issues here in the philosophy of so-called animal cognition or comparative ethology, but it seems to me that the burden of proof is with anyone who says no. The same issue arises, clearly, for human beings. So if we say that we do not know that the beagle feels love when he wags his tail and bays a bit and licks us and even gives us little nips behind the ears, and is obviously happy - more than happy - to see us, and delights in our presence, why would we not say the same about the human being doing these things, or their non-beagle equivalents? It's no good saying that he's doing it because we feed him. The same is true in the human case, but the manner of feeding is different, as is what is fed. It is difficult to imagine an ant loving us, but I think that is because there is no demonstration of affection from ants, no licking or running round in circles and so on. They would be ignoring us, if they were human and doing what they do. None of this is an assumption, though; it seems to be more of a common sense observation, but one that ignores false philosophical paths.
You have described a fascinating phenomenon that I think is remarkably common, though I don't agree that it always happens It certainly happens frequently in my experience. Perhaps we both have very bright colleagues whom we happen to know very well, and can anticipate what they will say! I am delighted to see "the phenomenon" so well described. However, in the form you present it, I think most philosophers and psychologists would say that the question you ask is a psychological one, not a philosophical one, and that no doubt it is amenable to empirical research. Still, it does prompt a philosophical thought or two. I am put in mind of Wittgenstein's observation that 'In philosophy it is not enough to learn in every case what is to be said about a subject, but also how one must speak about it. We are always having to begin by learning the method of tackling it.' Perhaps when you ask a colleague about your problem, you have to decide not just what to say but how to say it, and that is enough for your mind to turn up the answer. This is true in philosophy, but perhaps if the problem is one whose shape is completely obscure to you, it is in effect a philosophical problem for you, at least temporarily. Then you see the way through the confusion to what is actually going on. I wonder what sort of experiments a psychologist might suggest to answer your question, 'What principle could explain the phenomenon?' I am sure there are bad philosophical principles that some people would drag in, such as this: you already knew the answer in a previous life (Plato). I would go very gingerly with answers like that.