Having a submarine "train" for logistics strikes me as a bad idea either from an economic or military point of view.
The economic point, if there is any, in devising a submarine freighter is mainly to achieve lower drag by getting well below the surface, where wave drag on the interface is by far the worst drag surface ships have to deal with. Aside from the sheer novelty requiring an up-front investment in design, and the expense of a pressure hull, there are a couple of major drawbacks--one, you need to get the overall density of the ship up to that of water, which is much tighter packing than any surface ship. Obviously, since ships are meant to float! But the difference in density, that is the ratio of useful volume total to what is below the water line, is quite high, something like a factor of 4 or 5. There aren't a lot of cargoes that lend themselves to such tight packing! The sub will wind up carrying heavy metal ballast which only adds to maneuvering problems and offsets some of the advantage of avoiding surface waves.
Second, you need power, and aside from nuclear power, for practical purposes this means airbreathing--a snorkel that can allow the sub to remain deep enough to avoid surface waves would be a very long and draggy one. So, post WWII such a scheme would strongly favor nuclear power of course, and aside from a class of Soviet/Russian icebreakers, for various reasons the few commercial nuclear ships that have been commissioned have not stayed in service long. For the subs the necessity is obviously a stronger argument--but the question is, why have such subs at all?
Now--breaking up the load into modular towed pods seems extremely questionable to me. It certainly does allow the "locomotive" sub to be more agile on its own and removes the problem of making its density match the water--but only to transfer that latter problem to the cargo pods, with a vengeance! The favorable reduction in drag, which is the one thing this submerged cargo concept has going for it, is now offset or reversed by breaking the load up into multiple pods, each of which has its own drag and which probably interfere with each other. So economically it's a no-go--if you have a need for such subs, such as for hauling oil for instance, you probably know the standard cargo volume/mass to design for, and it will always work better economically to consolidate it into one well-streamlined hull.
I haven't addressed the military side yet, but in known history that is the one case where logistic subs have in fact been used. The Germans did, per a suggestion up thread, use submarine cargo ships in WWI, to trade with the USA past the British blockade. All major naval powers of WWII did use subs for communication with and limited resupply of various bases, and for special operations (but that's getting away from the cargo concept!) The additional concern military logistics brings is stealth rather than economical operations. (And, aside from scientific exploration, military stealth is in fact the only reason submarines have been used for any purpose hitherto).
There too the multiple-pods-pulled-or-powered-by-a-locomotive sub concept falls flat. Just as multiple pods will generate more drag, they must also generate more noise. Again if we are going to do this at all, it's best to do it in one integral hull!
The sub tug concept does have the merit of being flexible; you can haul a few or up to the tug's limit, many, modules. So if there were peaceful operations where there is a need to haul cargo to or from some submerged destination, it might be viable there. But not, I think, for point-to-point hauling from one surface port to another. Or only to achieve some odd purpose, like being able to haul a load under the polar icecap. Because it won't be economic if any more conventional option is open.