Without compression rods and plates, the cells within their uncompressed case will have somewhat shorter lifespan. But it might not matter very much.
With two parallel battery packs, each should be protected by
350A class-T fasting acting fuses on their way to a common "+12V" shared bus bar. Do not use a mere ANL fuses to protect BMS and the battery packs, they're far too slow to respond to a short circuit. This one, in quantity 2:
https://www.ebay.com/itm/284848435149. Or you could "downgrade" to smaller size (300A) and get a more famous manufacturer, with fuse block and cover supplied, here:
https://www.ebay.com/itm/265680106840
Photo show my own Bussman Class-T, with the containing project box temporarily opened and set on top for the photo. (Bus bars were also temporarily uninsulated, for the purpose of testing high-current "active balancing".)
For the case of Littlefuse, you will have the hassle of building or buying similar small project boxes to contain the fuses, connecting your top-most cells using an extremely short wire segment no smaller than AWG 2/0. Depending on the distance to your inverter lugs, you might want to upgrade the longer wires (from the "output" sides of the two fuses, each reaching to the Inverter's 12V input terminal lug) up to AWG 4/0, for less voltage drop while running at high power.
You might also want to add a coulomb counter for each battery (along the grounding cables into the Inverter grounding lug, although the BMS voltage readings should be your main guide regarding battery State-of-Charge. JK BMS provides a "remaining battery percentage" which isn't very accurate.
I use the Inverter's DC lugs as my DC bus bars. If you do the same, you will have 3 high quality terminal lugs stacked on each of the Inverter's DC power lugs (one from each battery pack, the other supporting TM shared 12v loads and grounding) In my case, the "12v" wire contains a smaller 100A fuse (not yet fast-acting Class-T, but i have ordered an upgrade for that ANL fuse) and ends at a plastic-enclosed "12v" power distribution block.
The power distribution block contains wires for the Solar Charge Controller, a fan for the battery+inverter compartment, an external 12v power port, and the WFCO 12v "main wire", which ends at another distribution block behinds the WFCO. (That 'behind-the-WFCO' block also connects the 120-VAC power Converter and two wires into the WFCO's 12-Volt fuse board, using both the "Inverter Input" and "Battery Connection" ports as input power connections to the Board. My Power Converter can exceed the 40 limit of the WFCO fused "inverter input" connector. Use of another power distribution block allows high current from the Converter to go directly to the batteries, bypassing the fuse board entirely. The presence of dual 40A inputs also allows the fuse board to receive more total current, although I don't think that such high current has ever been used by my "downstream" DC connections.)
Your JK BMS units don't have the heater function, but are otherwise similar to mine. Lot of settings will need to be tuned (especially maximum charge current, and over-voltage/under-voltage limits). The 'Discharge Overcurrent Protection should be pretty severely reduced, because JK does not have a separate setting for the maximum instantaneous current. (I've got my value reduced to 2 seconds, allowing for just a bit of peak draw from the Inverter on AC rotor start-up and possibly high-reactive loads coming into the Inverter from the Microwave Oven).
The default values of Under-Temperature Protection for charging, and its corresponding recovery parameter, are both FATALLY LOW for LFP battery cells. They must be raised above zero, "recovery" temperature raised first. My values are 3.0 C for charging shutdown, and 7.0 C for subsequent recovery.
The maximum CMOS temperature limit in JK BMS is not tunable, and preset to a value of 100.0 C. I feel that to be "too high" and wish that it could be configured to kill high-current discharge in cases where the measured temperature exceeds 85 C.
