Why are my batteries absorbing and floating at a temperature much lower than what is set?

Thanks for asking.  Yes, largely . . . slowly, slowly, I work my way up the learning curve.  I did post recently on my other topic.  I now DO understand even my original query about pv<load during the day.  bottom line on that was that a minor deficit is no problem, but I had to be sure my battery bank was big enough that if the sun went behind a cloud and most of the load hit the batteries it would not overwhelm them.  I am, gradually, getting reasonably comfortable, but there is a LOT of things that really don't have consistent answers.  Sometimes there are just too many variables, and often different manuals or different people have different ideas. 

For example, the Outback Mate3 manual tells us that (with some load, presumably modest) the red battery status light comes on when the batteries are below 60% or 45.6 V.  The energy cell manual tells us that if the batteries recover to 48.8V after 24 hrs rest, that represents a 50% SOC.   From experience, I would say there is no way a battery that is in 45.6 V or less while in use will recover, ever, to 48.8V.  Only if the load was really large. 

Meanwhile, also, the mate3 manual states that the green battery status light represents 80%+ SOC.   However, as far as I can tell, it is tied to the non-temperature compensated voltage, and it seems like when my batteries are fairly warm (31-34 C) I spend virtually no time at all in the evening before the batteries go to yellow (49.6) - within 10 minutes, generally.  I think this is just temperature related, that just as 60 is the new 50, so 49.6 is the new 50.4 V. 

Anyhow it makes it hard to keep up with things, if I have the system set to go to grid at 48V, then if the night is hot that can happen way too soon, before I have used anything like the amp hours that I would have wished.  It is all a moving target.  I would have loved to have put in a FNDC, but they local company tried to install it twice, and both times the hub stopped working as a result, so we agreed to give up after replacing two hubs.  But, I really hope I am getting a bit closer to some version of reality.

Another thing, I did an open circuit test of the batteries recently and the idea was to see if they were holding charge after 24 hours at rest.  However, according to the manual 50.4 V at the end of the game would mean the batteries are kaput, and 51.36 means they are great, but this is supposed to be temperature compensated and it is not clear whether that is the temperature at the start of the test or the end of the 24 hours. The average for each string was about 51.1 V, but the temperature at the beginning was about 34 C and at the end was about 29 C, either way it would make the corrected value a bit high.  Also, I have two volt meters and they give me totals about 0.2 V different one from the other.  The 51.1 V was on the lower one, the other would have the uncorrected value at 51.3 V, making corrected values even higher. The manual for them says they can have an error of +/- 1%. which at 50 V is +/-0.5V.  The individual batteries at the end of the test varied by 0.1 V on one string, and virtually nothing on the other string.  The two strings were the same though. Are the batteries on the one string that were lower a bit worse, and the ones that were better a bit better, or is it the total that counts?  They were similarly different at the beginning of the test.  It feels like getting lost in the details, but at the same time when the difference between dying batteries and healthy ones is just 1 V the details seem to matter. 

Hope all this at least helps someone else who is feeling overwhelmed feel a bit less alone.

cheers.