You could try 0.5 at 40W, would be interesting. Amp draw just under 9 so the MJ1 would work, just. You could do a paper calculation and make allowances for battery loads etc but real world tests are always more reliable.
Are you using a multimeter to check batteries and if so are all batteries keeping more or less a full charge after rest?
Could get boring after a while though
Measurements like these are good enough without being obsessive because they're meaningful to your equipment.
I may get round to trying something similar with my Pico mods though I'm satisfied the 30Qs give me longest runs.
In fact I could use 18650s in my Presa 100s which could be interesting.