my intuition is that the amounts of energy you could actually pull out of a system like this are not terribly high, on a per each/per size level, and the sea is pretty hostile to most available forms of technology (particularly when you're dealing with sea + volcanic outgassing). Given the rather severe difficulties in getting down there for repairs (remembering that the initial thought was "well, if we didn't *have* to make it human-acessible, it would be *cheap*) you'd have to have a system that was ridiculously reliable, while submerged in seawater, straddling a volcanic vent. I'm not saying that it's not doable, because I don't know enough about current state-of-the-art in such things to know for sure, but it would take a *lot* of initial investment to get the tech working properly, and I could see it being the sort of situation where discovering that you had a .1% failure rate rather than .001% meant that you went from being somewhat profitable over time to horribly unprofitable.
Side note: unless you have someone with a lot of money who is doing this for other reasons, you not only have to do slightly better than break even, you have to be able to give the investors a degree of return on their investment. Monetarily, if you're making back 2% per year on initial time and materials, that's hurting you.
Now, in a gundam-style universe, it might well make sense. In terms of current real-world? I'm suspecting not.
Side note: unless you have someone with a lot of money who is doing this for other reasons, you not only have to do slightly better than break even, you have to be able to give the investors a degree of return on their investment. Monetarily, if you're making back 2% per year on initial time and materials, that's hurting you.
Now, in a gundam-style universe, it might well make sense. In terms of current real-world? I'm suspecting not.