Concentrated solar thermal gets much better performance than nuclear thermal because you're only limited by the materials properties of your heat exchanger, not your reactor fuel. The problem with solar thermal is three fold.
First, you have to be able to see the sun, so your burns are restricted to times where you're not in Earth's shadow. That could make transfer windows difficult to hit, of course if you have a thousand seconds or higher ISP, that's not such an issue.
Second, we're talking about high thrust rockets so you don't have astronauts sitting in the radiation belts for days as a low thrust rocket spirals out. That means you're going to have to have significant structure to support that concentrator, and it's no longer going to be very lightweight. Of course with such a high ISP, this may no longer be such an issue.
Third, this reflector cannot simply be static. It needs to be able to articulate to whatever angle you need to thrust, which greatly increases the complexity and mass of such a system.
But if you want to move great masses you'll need something more powerful.
No one here has actually made an argument about why they can't be scaled up, but just waved their hands and said "Nope."
No one has ever said they can't be scaled up. The problem is one of electrical power. ISP is directly proportional to exhaust velocity. While high ISP means high propellant efficiency, it also means low power efficiency. 200kW fed into a moderately high ISP NTR is going to offer much more thrust than a very high ISP ion drive. On top of that, the NTR uses the thermal output of the reactor directly, while the ion drive is going to suffer huge losses first converting that to electrical.
You've still got a fairly expensive asset sitting idle.
Perhaps, but look outside any active distribution center and you'll find acres of idle trailers. I know Walmart even uses them as long term storage.
Then there's the extra labour to move one box at a time
If you're already moving one box at a time, because you're building the pallet, you may as well move one box at a time directly into the container. That was the whole point I was trying to make. There is no extra labor. The container itself is the unit of transport, rather than the pallet.
with ladders. You might be eight feet tall, but most people aren't.
Large molded plastic step stools, actually.
... two trips over one,
Unless the previous trailer just waits around until the next one is dropped off.
And where is this happening?
Anywhere you're shipping freight between two of your own facilities.
You can unload 3,000 cases of palletised goods far faster than two or three people stuck in the back of a semi-trailer can put them on a conveyer belt. It takes a warehouse 20 minutes to unload a full 26 pallet load semi-trailer, it takes 2hrs to unload 1500 cases on a conveyer belt.
That's only relevant if the driver and truck has to stay with the trailer. In a major distribution network where it's the same company on both ends, that does not apply.
Sure, pallets take some vertical space, but the amount of space taken is small compared to the ability for one person to move close to two tons of cargo single-handedly across smooth floors with no more than a jack.
That requires one person and one hand truck or fork lift to shuttle each pallet of two tons of cargo back and forth. It's cheap, but it's very low throughput, and it doesn't scale up to high throughput well. If you need high throughput, you're going to use conveyor. You use pallets for storage, for items that cannot be conveyed, or for things like LTL receiving where you can't keep the trailer, and don't want to hold up the driver while you hand unload.