The above example shows 8V sag due to 0.32 ohms battery resistance. This seems high and is one source of energy loss. How was this figure arrived at?
The controller in the above example is shown as a 25A controller, this might be a 6 FET which has more loss than a 12 FET that most folks use at 100 volts.
Calibration of speedometer is also important - differences in speed make quite a difference in efficiency. Also shunt calibration. A few amps of error makes a difference in the power used.
Altitude, air temperature, rider size, clothing and riding posture change the air drag. The simulator doesn't indicate what posture is in use, but that makes a significant difference at 37 mph.
Gradient is also important - a road that appears level may in fact be slightly downhill leading to error in the energy used.
Another difficulty is measuring steady state. Slight acceleration or deceleration is hard to detect and makes big differences in power used.