This has been rumored for some time on various railfan groups across the web and I've always figured it to be true. In a few instances, people have called out that trains are moving too quickly up grades for their supposed load or are way underpowered for a train if all the containers were full. It's all been a bit suspect.

This has only fueled a lot of skepticism about the BRI. There's a few primary reasons for this:

1. Break of gauge: Any of the current routes require gauge changes between 5' and 4'8.5". Each time this occurs it incurs either (a) translating costs to move containers from railcar to railcar, or (b) bogie change costs. This both slows it down and adds significant labor expense to the endeavor which reduces competitiveness vs sea. Talgo (in Spain) makes gauge-changing bogies which theoretically can handle this, but they are not designed for the tonnage nor are they particularly well-suited to the extreme conditions found on the routes.

2. Axle Loading & Loading Gauge: The axle loads across the different networks are highly variable, requiring cars only be as heavy as the lightest network they travel on. In addition, the loading gauges are much tighter in Europe, resulting in only single containers of shorter length being allowed on COFC trains (Container on Flat Car)

This isn't to say the problems aren't surmountable, just that they have not yet been and it is going to be difficult to do so in the future.