Draw down the battery, to ensure all power available can be used.
In the off-chance it decided to draw from one string, not the other.
All things being equal, maybe connections aren't equal, and something is heating up.
Or a panel is bad (although brand new; I would have suspected damaged one if used.)
Is anything running hot? MC4 connectors, wires?
Insulated parts you could test by feel. IR thermometer can check temperatures. IR cameras can show hot spots, and can show face of PV panel array.
If you had module-level power electronics, you could read them, see if any panel had low voltage (all in a string will have same current.)
Shutting off everything, or disconnecting PV if there is a switch (so no current flowing), disconnect/unplug one PV string and test one string at a time, see if they give different results. Swap PV string to other input.
Double-check number of panels in series, see if you connected more to one string than the other, or shorted some in a loop and left fewer in the string.
Although, none of your numbers indicated different number of PV panels with both strings operating at same MPPT (would have been same current, different voltage, unless Vmp was below MPPT minimum.)
Beyond that, you would have to test individual panels or divide and conquer with shorter strings to find the odd panel. Tedious to make and break so may connections. Always interrupt current before unplugging.
Individual panels can be tested for Voc and Isc. That alone doesn't catch all bad panels. A PV panel tester will measure Vmp and Imp. I used a resistive load (space heater) and found a couple bad (17 year old) panels with different V(load) and I(load) from the rest.