Because many ships spend more time not performing the task they are designed to do and more time in transit.
This is true.
Thus objectively being more efficient use of materials by having them be "paused" while "docked" in the container ship.
This, however, is rather not, at least for the only application we've been discussing in this thread. Each kiloton allocated to hangars on a container ship only gives you 625 tons of containers, which then incur further loss from the structure and crew requirements of the containers.
Let me give you a real world example. All my light transport ships are built on the same frame. They all displace 100Kt, go at 1000km/s with my current tech, and aim for roughly the same range.
Chungking-C Colony Ship: 77.5Kt of Cryogenic Transport for 310K meat popsicles.
Ghent-B Troop Transport: 75Kt of Troop Transport Bay for 75K in troops carried.
Aberdeen-D Freighter: 75K of Cargo Hold for 75K cargo capacity.
Wenchow Container Ship: 72Kt of Commercial Hangar deck for 45K capacity for containers.
Containers are built as close to 5k as possible.
Cryo loadout: 43.2Kt (9x4.8Kt) of Cryogenic Transport for 172.8K meat popsicles. 55.7% of the dedicated ship.
Troop loadout: 43.2Kt (9x4.8Kt) of Troop Transport bay for 43.2K of troops. 57.6% of the dedicated ship.
Cargo loadout: 40.5Kt (9x4.5Kt) of Cargo Hold for 40.5Kt cargo carried. 54% of the dedicated ship.
In each case, the dedicated ship is much better at doing the one job it can do than the modular conveyor. (The fact that there's no hold smaller than 500t hurts the cargo variant quite a bit.)
And as far as minerals go, in tons of TN mineral per:
Wenchow, hull: 6596.9
Cryo Container: 316.3
Barracks Container: 137.3
Cargo Container: 38.7
Chungking-C: 4391.9
Ghent-B: 3036.9
Aberdeen-D: 1606.9
Wenchow + 9x Cryo Container: 9443.6
Wenchow + 9x Barracks Container: 7832.6
Wenchow + 9x Cargo Container: 6945.2
Wenchow + 9x each: 11027.6
A single Wenchow with the option to carry any of the three loadouts costs you 122% as much minerals as buying one of each dedicated ship.
So, let's look at those numbers a few different ways:
Wenchow Cryo vs Chungking-C:
2.15x cost per hull.
3.86x cost per colonist carried.
Wenchow Barracks vs Ghent-B:
2.28x cost per hull.
3.96x cost per ton of troops.
Wenchow Cargo vs Aberdeen-D:
4.3x cost per hull.
7.9x cost per ton of cargo.
Wenchow + all container loadouts vs Chunking-C+Ghent-B+Aberdeen-D:
1.22x cost of buying all three dedicated ships.
For ship types that are going to be running all the time, like your colonial transports, using Modular Conveyors is very clearly not worth it. For pure cargo hauling, even if it's going to sit idle a decent portion of the time, you're still probably better off building dedicated ships. But covering your needs for troopships/supply ships/colliers, there's clearly a place for Modular Conveyors, as you can meet your peak demand for those types of ship without having to keep a bunch of idle hulls around, or dedicate multiple shipyards to their production.
Purpose built ships give you efficiency. Modular ships give you flexibility. There's room for both in your merchant marine.
1k ton is the traditional size for containers in AuroraVB due to not needing a bridge.
This strikes me as very much a false economy, at least now under C#. (I dunno what particular ship design quirks under VB6 made it make sense to do that.)
The bigger you build your containers, the less efficiency you lose, though you're giving up flexibility. Building out a single container that fills the Modular Conveyor is the most efficient thing you can do, if you a) don't need to ever carry split loads and b) Don't need to ever use ships of a smaller hangar size as container ships.
A 45K Heavy Cargo container, for example, can dedicate 99.17% of it's displacement to cargo holds, compared to 97.15% for the 5K version. It's not going to shift the efficiency numbers in favor of modular over dedicated, but it's something to consider. You also, and I didn't even think about this until now, save a lot on mineral cost for the containers. Again, not enough to make a big dent in the mineral cost per ton of cargo carried, but it's a thing to think about.