It's definitely an electrical potential related problem, but not a classic ground loop like that with active speakers for example.it's extremely likely you just had a ground loop
The RME is an IEC Class II device with an external DC power supply. The problem occurs with USB devices and the mainboard audio (is that run via the USB controller?), however it is very variable in quantity. With the Topping E30 (DAC meant for desktop applications with seperate data and power lines) for example it makes about a 2dB difference from 110.x to 108.x, repeatable and definitely measurable, but that wouldn't constitute an audible concern. With the Apple dongle the effect is far, far greater.
With the dongle run from the laptop that is then plugged in via its power supply (also Class II), a 50Hz spike (and its second harmonic) becomes visible in the measurements, but the noise floor doesn't really change (and yes, I should use averaging to investigate such effects).
I suspect the dongle is just very sensitive to any supply voltage irregularities - don't let the snake oil brigade read that though
With the perfectly clean DC from a battery powered device - the way it was intended to be used - it just simply works best.