I don't have enough data to produce a full report but after 4 months of working pretty much every day on this project, I finally just sealed up everything up last week and said if it burns the house down, so be it. I have dual 280aH banks wired in a mirror image, same exact length of cables and connectors and same dual side-by-side JBD 200A BMS. I have dual battery switches with dual same thickness solid copper buss bars feeding from a 2/0 battery cable through dual 200A bussmann fuse bars through dual 350A switches connected into my house 4/0 welding cable inverter feed. I honestly think this BMS must be a prototype version or something. I've looked at my cell data charts via Android and there is no way that these cells have peaked up as high as 4.2V and down as low as 2.5V and danced all over the spectrum like these graphs show. I never put the voltage on the house line for them to do that and there's no reason they would have ever had spikes down to 2.5V. Additionally I have gone through all the custom Android parm settings about 6 times, writing everything down and both of the identical 200A BMS configurations are exactly the same. And no I do not have loose buss bars because I have torqued them 3 times and I can run a 150A microwave for 8 minutes and all the cells measure the same and every buss bar is cool to the touch, so I think it is the BMS and not the installation. This is also proven by one BMS (while switched both banks on long-term in tandem) shows all the individual cell voltages as equal, yet one BMS says I am 35% power and the other says I am at 65% power while both read identical on the cell voltages within 0.02 (.) Also one BMS shows I have 62 ? "time left" and the other shows I have 158 ? "time left" - whatever that means and this is noted while all the cell voltages on both banks have been working in parallel for several days while all cells are showing the same voltage within 0.02 of each other. Other tidbits include that I wired in dual identical coulomb meters (battery monitors) with the shunts wired right before the BMS grounds to the frame so I can double check what the amp meters on the Android panel is showing. Watching the Android panel meter vs the battery monitor will show that if charging option is turned off, the meter on the Android panel will freeze while the battery monitor shows it is actually still discharging like you asked it to, which means the firmware is doing what you ask but the meter on the display stops showing discharge amps when the charge option is shut off. So I am guessing there is some bad firmware going on here or some other extreme oddity is causing the differences of reporting on the BMS Android panels because with the bank running long-term with the two tied in parralel, the voltages of all 8 cells stay very near the same but the BMS is all over the map as far as percentage charge, ???time-left??? and the operation of the meter with discharge/charge off/on is messed with.
Side note, when I switch off the charging option on both BMS panels, again with both banks in parralel, by simply turning on the microwave (150A) caused one of the BMS to automatically turn charging back on its own. So there are too many flaky things with these BMS units to try to set up any kind of scientific lab work or gain too much of an education about the workings of LiFePo4 in my opinion to invest too much time into it other than saying what his presented here is better than nothing at all.
Otherwise, way too much time, unexpected outflow of too much money for all the connectors, cable runs, etc, etc, etc that I am going to let this setup run now without wasting any more of my life on this setup and if something goes or burns out before 10 years, chances are I am going to be gone before they are and especially if I take on another project like this. This one is simply not fun any longer. Also if I cause one cell to last only 5 years instead of 10 because it let it float at 13.8v instead of 13.6v, it's now all so clear that in time and worry, it's cheaper to just buy a new set of batteries if you find their life has been shortened (at least for someone like me in the 12V market, as far as the homeowner with banks after banks of these 280 3.2A - all I can say to you is good luck brother managing these.
P.S. I forgot to note that in the tight quarters where I had my old lead/acid solar controllers mounted, this took a couple of weeks to remove and re-mount almost $300 worth of new solar controllers for my two dissimilar size solar arrays. So short story is this is great for someone who is looking for something to do but if you are aching to get back on the road again, it's one project that can quickly turn into a career rather than a single "winter project". Also note that if I was going to watch the meters and shut off charging when it got above 13.6 and discharge my cells to where they only got "this" high and discharged down to "that" low then I would tow a trailer of lead/acid before I'd spend any more time on this. My cells "go to sleep" anyway when they get to about 13.8V and I know the meters are accurate because while the Android meters for the 2 BMS show "0", the battery monitors show only about 0.05 to 0.1 charging amps when the BMS meters are showing zero. So since nothing trips off any of the BMS cut-off criteria, I assume once they reach this voltage level, they appear not to accept any more amps even if I start the alternator to pump some 14.2V into them. Either it has to be the cells don't take more current when they get to 13.8v (divited by 4) or otherwise and unless there is an entire other program executing in these BMS units that is not documented, they are "quite done" at 13.8V. Incidentally, when I have read of all the people saying that the alternator at 14.2V needs a B2B to charge at 14.4V, I have to wish I was on that influencer payroll. If you've asked me the very largest surprise and revelation of what I've learned with this project, it just how much pure BS is circulating about the way these cells work. Buenos Suerte! YMMV