I have parallel split phase 6500's. All of my batteries to bus bars are equal length. Positive and negative cables are the same length to the bus bars with 4/0. The positive cables from the buss bars to the inverters are the same, but one of the negatives is 2 inches shorter than the other. These are 2/0.
Does 2 inches matter? ............ (I expect some joke replies)
You're more likely to have variation in terminal lugs and clamping connections of lugs.
Fractions of milliohms difference matters. Inverters regulate their output AC voltage so more DC cable voltage drop means more battery current is required by inverter to produce the same AC output power with lower DC voltage input to inverter. Inverter conversion efficiency drops for lower DC voltage input.
The object on parallel battery hook up is to get balanced current. Some imbalance will be due to batteries themself due to mismatching of their terminal overpotential voltage slump vs. load current demand. Older LFP battery will likely have greater internal impedance causing it to contribute less current to the parallel combination of batteries.
Don't forget connectors and bus bars connection resistance. A poor bus bar connection can screw up the current sharing between batteries a lot.
It is tough to get better than +/-10% current matching. Better than +/-20% matching is fairly satisfactory.
If you use separate BMS's on each parallel battery array they can help with parallel battery current balancing. The greater the current through the BMS the hotter their series MOSFET switches get. The hotter the MOSFET's gets the greater their series resistance. For parallel batteries with their own BMS, if one battery hogs more of the current, the BMS series resistance will increase due to heating, lowering that battery's dominance of supply current.
This assumes the BMS's start with similar series resistance MOSFET's. Particularly, if you use different manufacturer's BMS's between batteries they are not likely going to be matched in series resistance.