The device is probably just using a USB-C format connector to get power, without using the data connection at all, and a strict implementation of the USB protocol on the other side (the so called Host) would mean the device gets from the host only the minimal power levels (100mA @ 5V, if I remember it correctly) meant to merelly power enough a connected device which has no batteries (say, a mouse) for it to actually do the initial USB connection negotiation, and that current will only get increased by the host it if during that negotiation the device tells the host it requires high-current (which in different USB versions has a different value - in USB 1.0 it was 500mA but latter versions increase it), a negotiation which that device can't do because it doesn't actually do USB data at all and just treats the whole thing as a dumb power cable.

Dumb charger bricks don't care at all because they themselves only do power and not the USB protocols, so really just treat the USB cable as a power cable into which they always make available whatever current the other side pulls up to the brick's max supply capacity (usually 1A or 2A) with no "USB negotiation".

This is why even in the times of USB-A some devices would charge fine from a dumb USB power brick but charge really slow if connected to a host which is a data device that can also do charging (like, for example, a notebook).

This is even without getting USB PD into the mix.

Because USB PD is a comms and power protocol, were the device tells the host the characteristics of the power it expects to get (not only current but even voltage) the USB PD brick has a proper USB implementation were it acts as a USB host.

I expect the USB PD brick has a strict implementation of the USB protocol which, in the absence of USB negotiation, just provides that minimum current that per the protocol a USB host is expected to provide pre-negotiation, which is too low for properly charging most things.