I haven’t had much luck getting to 0.5mA with mega1284, it was with the Mayfly 0.5b- https://www.envirodiy.org/topic/testing-power-consumption/#post-15252 My notes on measurements is that it was between 2-2.5mA. I did get well beneath 0.5mA on another project with mega2560, but that was a lot of hardwork, and also ended up with a higher value of mA for the production version – as the modem connection really consumes the energy.
In looking at testing the whole system, including production software – I came to the view its the apparent reliability of the overall system – and managing the power demand, and delivering readings to MMW, that results in a stable overall system – requiring less intervention and maintenance.
I put a model together for power used over a day with the most active subsystems. I enclose “Mayfly MS Energy Useage 231218.xlsx”
This is really paying off with this system https://monitormywatershed.org/sites/TUCA_GV01/