I agree with Robyn that it's important to consider the ethics separately for the cases of the individual animals and the wild population as a whole. Much of TONMO's hobbyist advice is aimed at encouraging ceph-keepers to be as well-educated and responsible for the well-being of their pets, and I think, unless we take the stand that they shouldn't be kept at all, that's something we're doing quite well.
It seems like the other case, the impact of collection on the wild population, is a whole other kettle of fish, so to speak. There are a lot of components to this being tossed out in this discussion, ranging from economics to transportation logistics to collection priorities to importing to regulation.... ultimately, these are probably the main areas in which we could affect change somehow, although how is not clear.
Robyn touched on something that seems key, though: the impact that collection has on the wild population depends heavily on the life cycle of the animal. It also, of course, depends on a lot of factors in the wild, such as its habitat and how easy it is for collectors to capture a large fraction of the population. Some animals are so easy to collect, because their natural defenses are not at all well suited to avoiding human assault, that they can be over-collected with ease-- I put seahorses in this category, since their main defenses are to hide among plants and to have armor and generally not be a good meal, but neither of these protect against major netting efforts. (When they're to be sold dried as a medicinal product, it's even easier to collect them).
In the case of cephs, it's rather hard to assess the impact of collection on the adult population. Coleoids seems to generally share the reproductive strategy of producing very large clutches of eggs and (para)larvae, but to have relatively little chance of each baby reaching sexual maturity and reproducing. I'm not sure how to assign even back-of-the-envelope numbers to this: it means that the population can theoretically bounce back rapidly from a small population to a large one, particularly since they have a short lifetime to reproductive age. However, it also means that removing adults or near-adults from the breeding population can have a very rapid impact on the reproductive capacity of the whole species, and also narrows the gene pool. One thing I'm pretty sure it means is that collection of hatchlings is much less likely to have an impact on the wild population than collection of adults, since the survival rate for a hatchling is very low anyway... it also arguably(?) offsets the cruelty of pulling the animal from its natural environment when its natural environment is full of nasty hazards, so it would give it more of a chance at a long and non-traumatic life.
I don't know that it's at all practical to somehow encourage collectors to collect paralarvae instead of adults, and of course they'd have to be raised for a while before sale, but I bet they're easier to ship.
In a completely tangential thought, I wonder if this reproductive strategy was part of why Coleoids survived mass extinctions that wiped out all other ceph populations, in that a few octopus-ancestors that had some particular mutation that helped handle the environmental conditions leading to the mass extinction could have expanded to fill niches left by other animals that weren't able to handle the changes.
Anyway, I know Eric Hochberg advised the California Fish & Game regulators to forbid bimac collection for the hobby trade because he believed that it would be easy for heavy collection to wipe out the breeding population of bimacs in the wild. I infer, at least, that a lot of you with experience in this area are concerned about that in the case of the "zebras" as well. Of course, this also ties into the discussion elsewhere about how overfishing of terminal-spawning squids is likely to impact their populations differently than overfishing of animals that have different reproductive strategies and timelines.
I've noticed, though, that several of the other animals that have been brought up recently (particularly Tasmanian Tigers and Hog Island Boas) seem like they have very different life and reproductive cycles than the cephs... Fish are closer in some ways, but I don't know enough about their details. Assessing how the animal's actual lifecycle will interact with various pressures, whether they be overfishing/overcollection, habitat destruction, water temperature changes, invasive species, or any of a number of other things is a very complicated problem. Of course, it's generally better to err on the side of caution, since it's not always easy to tell what impact these things are having on a population until it's an irreversible problem, but it seems worthwhile to look for as many answers as possible along the way, lest the conservation strategy turn out to be a case of "the cure is worse than the disease," or just putting a lot of effort into addressing one part of the problem, when the highest impact on the species' survival might be in some other area.
from me, is that sufficient to maintain my reputation for "walls of text," Dan? (I considered trying out e e cummings formatting for this post, but old habits die hard)