Yes, as listed in ADC0CON Bit 10 ADC0CODE the "binary" is two's complement coding.
And as can be seen from the code provided:
Voltage = ADC0DAT * Vref / Gain / (2^N-1) for uni-polar coding
and as can be expected for tow's complement
Voltage = ADC0DAT * Vref / Gain / (2^(N-1) - 1)
Simply try it by modifying the appropriate line in the code.
Thanks for that hint - we add with the next update of the data-sheet to make this a bit more obvious.