Rotating the cells of the pack when in a series serves no purpose like it would if they were in a parallel, if they work best in a particular order you might consider installing them that way every time.
https://redirect.viglink.com/?format=go&jsonp=vglnk_148359602427212&key=ccc2217bc2b75c4bdf5f7e57267f0464&libId=ixjyq4sw010009xy000DLqiyi2ag0&loc=https%3A%2F%2Fevolvapor.forumchitchat.com%2Fpost%2Fearly-firmware-7508618%3Fpid%3D1293056071&v=1&out=https%3A%2F%2Fdownloads.evolvapor.com%2FSetupES250.exe&ref=https%3A%2F%2Fevolvapor.forumchitchat.com%2F%3Fforum%3D454151&title=Early%20Firmware%20-%20Evolv%20DNA%20Forum&txt=https%3A%2F%2Fdownloads.evolvapor.com%2FSetupES250.exe Darn that's a lot of text, must be a better way...........but it does work.
If you feel like affixing a jumper wire from the output and ground for the 510 on the board and use atomizer analyzer in Escribe that will narrow down to whether it's a problem with the board or not. Kind of extreme perhaps but effective.
It would pertain to all cell types. To be more clear perhaps, two (same) cells whether connected in a series or parallel would have the same watt hours when paired. https://milliamps-watts.appspot.com/
Escribe will show what the voltage of the cells are under load. It sounds as if the board is not getting good power. Might be the batteries, might be bad connection(ect.) of battery contacts to battery. Might be bad solder joint somewhere between batteries and board.
Where you bought it if new, Lost Vape service center or a ticket with Evolv. 4.2 volts is full charge for most Li-on cells, there are some that are 4.35 volts.
The curve can probably be found on this forum. I calculate 12.5% less than stated mah for 18650's ( on a 200 board) Keep in mind that only the "fuel gauge" is effected not battery performance with the csv