Sure, I'm a model railroad guy, and I know you can look at freight cars and tell what the density of their targeted load is by how small the car is.
Little two-bay covered hoppers and ore cars are very short, 20 to 26 feet or so and much shorter then say grain or plastic pellet hoppers or coal gondolas, because the density of heavy frac sand or iron is much heavier and in the US typically goes to 4-axles so those weights cannot exceed the tons per-axle loading limit of the track.
I also know that shipping out things like animal feed and wood on containers often does not 'top out', because they are too heavy load bearing for the containers' lighter construction and can shift, which for a truck or a ship could be particularly devastating.
On the other hand our Traveller ships have a LOT of power in both thrust and computing, materials technology we can only dream about now, 1000s of years of interstellar commerce, and above all a need to make money from shipping.
They are going to solve higher loadings then what we can do presently, because the money says they need to.
So yes to limits of density per dton, no IMO to the limits you propose.
Consider just something like a basic ore- given your limits we are talking maybe half a meter tall worth of ore bin per dton, 2 metric tons per dton.
An average iron ore car (nowadays carrying taconite) is something like 90 tons loaded, ore and car, 50-75 tons of ore, and thus putting 23 tons per axle, most track is now set to handle heavier loads, to 125 tons total and 30 tons plus per axle.
https://www.american-rails.com/jennies.html
The iron ore car would work out to say 23 feet by 9.5 by 10 feet for support frame and carrying hopper, minus the trucks but likely covering for transport. Let's call it 80 US tons.
That translates to 2185 cubic feet, 61.87 cubic meters so let's call it 4.5 dtons.
80 tons works out to 72.5 metric tons. Dividing 72.5 by 4.5 gets us 16 metric tons per dton- over the proposed 'water limit' but not by that much.
Given that we are talking about ships that are flying AND may be at quite steep angles to planetary gravity at times, I don't think I can reasonably claim that they would have as much stability as a rail car.
So dropping per ton per dton down by a 1/3 to 10 metric tons per dton gets us ore that only fills up about halfway up the ore car form factor- but is very comfortably within TL5 to handle for containment, and I think reasonable although approaching limits for our higher tech.
Note that a goodly portion of the weight is 'tare weight' for the ore jenny. A smart shipper is going to look to use material technology to lower the container weight and therefore be able to ship more paying cargo per dton/ton limits.
Of course the free trader captain needs to inspect those containers to make sure shoddy ones aren't loaded that will burst and let free-roaming cargo damage the hold and contents.
As to liquid shipment, of course the same principles apply, but even so the standard 5,000 gallon tank cars of old have yielded in progression to 8,000, 11,000 and now 31,000 gallon cars. Track loading and materials allowing for larger and thus heavier tank cars to be used, all for the same purpose- more paying cargo shipped per car, less tare weight.
I would think perhaps there is a TL progression to be had here, as higher tech ships allow for greater load bearing, gravitic safety measures, smarter reactive maneuver programs,etc.
Just as a starting point for discussion, perhaps 5 metric ton per dton as you suggest at TL9, and then add 1 metric ton per TL. By TL14 you can load my 10 metric ton limit, and TL15 would be 11 metric tons per dton.
This should allow for multiples of the standard shipping fees and provide a payoff for operating those higher tech vehicles. Of course, the shipping cargo has to support the standard too, and so maybe the majority of free traders have to work the lower dton standard because their customers do not ship in TL14 containers.
