I’m not sure why it’s higher for XLR. Effect of the shield on overall capacitance should be lower for "floating" (not GND related) outputs (like microphone or transformer). This should reduce overall capacitance. Geometries might be a little different, so perhaps this can affect capacitance (conductors closer). Wires in XLR cables are twisted and twisting increases capacitance slightly. Also, I remember similar effect with Benchmark DAC1 output. It had nothing to do with the cable itself but with internal gain setting. Output impedance at -10dB was 1600 ohm, and only 60 ohm for 0dB. I used it at 0dB position with narrow, uncomfortable volume pot adjustment, but sound was more "vivid".