BLE.on actually reduces Argon power usage?

These Argon power measurements seem a little odd to me. Anybody with thoughts as to why this might be the case.

OS: 5.1

Code that does nothing, radios off. - Average 4.85 uAmps measured over 60 seconds
Code that just turns on the BLE radio - Average 4.6 uAmps measured over 60 seconds
Code that turns on BLE and connects as a UART to my phone - Average 4.85 uAmps measured over 60 seconds.

Why would just turning on the BLE radio reduce overall power usage?