Could also be due to to long/to thin wiring from the battery to ESC.
For example, by the calculator if you have 12AWG wiring in total of 30cm (each wire 15cm), the resistance is around 0,003 ohm.
Then let's say on the connectors we have 0,001 ohm in total (just an assumption)
Internal resistance of the battery 0,003 ohm for each cell,so for a 3S Lipo, this gives us 0,009ohm for the pack.
Let's add a 0,001ohm for the ESC itself (depends on where on the PCB the actual measurement is done)
So a total of 0,014ohm. Doesn't seem much? Now introduce a 100A current spike in there and you get a voltage drop of 1,4V in total on the ESC voltage measurment. So if we have a 3S battery, the ESC sees that as 0,47V per cell.
Of course the current draw is much less in average, BUT if the ESC is not done properly (not averaging the measurement but instead just trips on the voltage drop) then yes, this can be a problem....
So it would be good if you could provide some more information, like:
-cross section of the wiring (in AWG or mm2)
-length of wiring
-internal resistance of the battery
-are connectors OK (how warm are they after the run?)