Turns out it's pretty tricky to make a padlock secure against bolt cutters. Of course improving the security of the lock would soon lead to thieves cutting off the locking bars on the container instead.
It doesn't have to be the lock that's secured. What about a hinge mechanism which takes a programmable amount of time to open the container door? If it takes 10 minutes for the door to open, the lock doesn't matter as much, since the thieves have a very limited amount of time before the trains start moving again, apparently.
Those are standardized containers standardized which are made in the millions as cheap as possible. That stuff needs to work, think of all the means of transport where it is used (vessels, trains, aircraft), be accessible for inspections (customs) and so on. The costs to secure and upgrade the containers for the whole system is probably a lot higher than accepting a small percentage of stolen parcel.
That sounds like it could be quite error prone. If it's electronic, what about power failures? If mechanical, sounds like lots of intricate machined parts which will wear down over time.
Time locks have been around for more than 100 years, I'm pretty confident that if something like this was desired by the powers that be, it would have been implemented. What I'm assuming is stopping it aren't the engineering challenges, but the fact that currently it's cheaper to let the thefts happen than it is to mitigate them.
I wonder what the cost of engineering secure containers would be vs. addressing the societal problems in the USA that lead to situations like this would be.