@Wawa identified a number of problems using a ACS712 with the unstable reference voltage of the UNO etc
ran a test using your code of post #1 with a couple of lines added to print the power supply voltage
Serial.print(" Supply Voltage = ");
Serial.println(5/1023.0*analogRead(A1),2);a run for a few seconds (power supply voltage 2.23 current 68mA)
11:56:41.074 -> Voltage = 2.49 V Current = 77.07 mA Supply Voltage = 2.24
11:56:42.409 -> Voltage = 2.49 V Current = 76.88 mA Supply Voltage = 2.24
11:56:43.775 -> Voltage = 2.49 V Current = 79.35 mA Supply Voltage = 2.23
11:56:45.141 -> Voltage = 2.49 V Current = 78.90 mA Supply Voltage = 2.23
11:56:46.469 -> Voltage = 2.49 V Current = 79.40 mA Supply Voltage = 2.23
11:56:47.828 -> Voltage = 2.49 V Current = 78.25 mA Supply Voltage = 2.23
11:56:49.193 -> Voltage = 2.49 V Current = 76.98 mA Supply Voltage = 2.23
11:56:50.525 -> Voltage = 2.49 V Current = 76.67 mA Supply Voltage = 2.23
11:56:51.886 -> Voltage = 2.49 V Current = 78.08 mA Supply Voltage = 2.23
11:56:53.239 -> Voltage = 2.49 V Current = 79.79 mA Supply Voltage = 2.22
11:56:54.605 -> Voltage = 2.49 V Current = 78.35 mA Supply Voltage = 2.22
11:56:55.960 -> Voltage = 2.49 V Current = 76.91 mA Supply Voltage = 2.23
some sample results taken over a range of voltages
meter UNO
V mA mA V
0 0 7 0
.59 12 18 0.57
1.55 40 48 1.51
2.56 64 74 2.52
3.57 92 103 3.51
4.76 120 129 4.38
6.98 184 197
10.2 275 290
15.9 417 430you could do a calibration run to read the UNO measured current with 0 current flowing (7 in the above result) and subtract it future readings
however, if you require milliamp measurement with two or three places of decimals a better measurement device is required
I did use a ACS712 some years ago to measure the current taken by an extruder ram - when the current reached a set level (indicating ram was reaching the end of the extruder) the ram would be stopped and withdrawn - however, the current measured was several amps and great accuracy was not required