Incorrect sensor readings HS110... Anyone have similar?

I noticed today that my watt reading (which is pretty accurate when checking against different devices) is different to the Amps*V calculation.

I’ve setup 3 Amp sensors:
Raw sensor,
Watts/Volts,
Raw Sensor with a 0.2 offset (it had a 0.2a difference at my PCs normal load.

See the album here: The calculated and the RAW output match when the load is high but not when it’s low.

Could TP Link use a curve fitting bias to the watt measurement (I guess the amps aren’t super accurate at low load???).