Padwan

This sounds like a PV "freeze"; others have experienced PV "dips". There is a long discussion on it here and elsewhere.
But first: is it possible that you have too high a PV voltage, perhaps from 3S of 72-cell panels? PV power will derate above 130 V, and is down to zero power at 145 W.
And second: is your float voltage setting (setting 27) lower than your bulk/absorb voltage (setting 26)? If so, when the battery is declared fully charged (it may not be, due to the premature float issue), then the charge current will be zero for a few minutes (up to about 10 minutes with light loads). But then it should keep charging to maintain the battery at the float voltage setting.
So third: what is the middle ("charge") LED doing when this happens? Is it off, solid on, or flashing?

That's easily arranged. Unless you are particular and want paying work....

Thank you @Youda.
This does makes sense, guessing the potential difference would cause voltage fluctuation between them and MPTT's would constantly be adjusting.... kinda like no win wars ...
 
oh and P.S swt setup you got. envious..

Because you are feeding the inverter from an E/L circuit in your grid DB, and you have connected N-E on the outputs which is an earth fault.
I hope the attached sketch is legible, my CAD skills are a touch lacking (I can do it but it will take the whole day!)

Ah, my bad. I took this as an ohms reading, with the inverter off and isolated. Odd that it's intermittent, though. Maybe it depends on a load between the DB and inverter, like the light bulb (circle with an X through it) on @plonkster's diagram. Perhaps your stove, oven, or geyser turning on an off? Or just leakage?
Some Axpert models only had the three relays; Weber's speculation is that those omit the inverter output relay. Since you have all single pole relays, and no setting 38, you can't do the proper switchover.
I can't see the harm in shorting AC in neutral to AC out neutral. Though doing so does short out one of the "safety" relays.

Possible violation of regulations, or at least norms. In South Africa, I seem to see a lot more double pole breakers/isolators than I do here in Australia. I have no idea about the SA regulations.

It is going to depend on what loads you have on there. Most old-style appliances are not going to affect this, but the new electronic stuff have stuff wired between live/neutral and earth and that tends to create little voltage dividers that raises earth roughly to the middle of the rail. Different loads... different voltage.
Because the neutral is now connected to the grid neutral, which is bonded to earth on the supply side.
That would be the easiest way to make it safe...
But there is that. Now the electricians I've spoken to over the years differ on this topic. Some see no problem with it, others say no. From my own understanding of SANS 10142-2-1, on single phase systems all disconnects (including the one in the inverter) must switch both poles, so it would not be legal.
There is however a diagram in the old SANS regulations that allows such sharing of a neutral on three phase systems in cases where the backup supply (that is the inverter) is connected at the "entry point", in other words, if the entire house is on the inverter. The moment you supply only a part of the installation, then you're supposed to have a double pole changeover, separate neutral bars, and so on and so forth.
I suppose one could always wire a small double-pole 12V relay in parallel to one of those grid switching relays (ie, solder wires onto terminals inside the inverter!) and use the NC side to power a bonding relay... It's a long shot...
For now you'd have to choose the lesser of two evils, which I suspect would actually be to bridge the neutral across.

I think that's 220nF capacitor. It has "224k" written on it. Those three relays are all single pole.

On the inside, you can look up the part numbers of the 4 relays at the bottom, eg bottom right:

If you look up the "Song Chuan 832ha" relay series, you will see that the 4 in this one are all single pole. If one of them are double-pole... then maybe. But then I also wouldn't expext the floating neutral voltage you noticed... so I kinda sorta think you have a model without it.

But for Pilotfish solution to work correctly you require a change over switch 

Take a look at Annex P.4 starting on page 361 of SANS 10412-1:2017 ed2
I sadly cant share this with you because my scanner has recently croaked and I haven't got around to sorting it out, but in summary you should bond your main E terminal (from earth bar in Main DB) to your inverter N terminal (at the inverter) which will prevent your floating voltage situation. The earth conductor should be a minimum of 50% size of the Live conductors.
This should be done before the inverter output Earth Leakage which I am sure you have installed.
So as an example if you run a 50amp 10mm L/N from the Main DB to your Inverter L/N(in), then you should run at least 6mm from your main E-bar to the inverter body and to the N(out) terminal (looped, not 2 separate wires).

Dang those inverter part numbers.

Unfortunately, the service manuals don't seem to be any help with troubleshooting the SCCs. I think all you can do is to check the MOSFETs and their drivers. Hopefully there will be enough drivers with similar components that you can compare one to the other. It will probably an interleaved configuration, like this:

That's about all I can find in the service manuals, sorry.

Since 51.8 V is the voltage where balancing starts, you might want a smidge more than 51.8 V (say 52.0 V) for the absorb voltage. Just so your're sure that balancing happens most days. On a totally cloudless day, an Axpert could exit the absorb stage a few tents of a volt under the setting. Even though most days it will overshoot, you can't rely on that 🙃

Those are the latest factory firmware versions for 5 kW (PF1) machines without the 64 V capability. Unfortunately, there is no patched firmware for these, to fix the problems that @Padwan has mentioned. If you happened to buy a 64 V model instead, it would have come with firmware version 72.20, for which patched firmware is available.

LOL, I hear you.
I'm certainly having fun its become my new "obsession". 
Having said that I do believe in buying and paying for good quality products that "Just Work TM". Thing is that I simply need more time to experiment and experience to define what is is that actually suits my needs.
Rationale here is that system im playing with now is going to serve as auxiliary backup supply, once I start understanding what my actually needs are and what is the premium product that Just Works TM, buying spares seems like a reasonable thing to do.
 
Until such time that I have option to have grid tied inverter [Currently not possible under Eskom, who is my provider, as i understand it] I figured I limit cost, play learn and have fun. 
Always open for guidance and experiences of others when it comes to their equipment and use cases.

SOLUTION:
So this was all part of confusion on my part. After reading the pylontech setup manual again,  i've changed ADD dip switches to 0000 on master and slave battery.
This change has enabled ICC to read both batteries stats, and looks all is in order now. 
Thank you @Chris Hobson  and @Manie

thx @Jaco de Jongh & @plonkster i guess whilst we await Manie to get a spare cycle to glance on this, ill get that box off Pi's and try a different one in case there is a genuine hw issue.