Pete.J
home --+--a---+--b----(meter)---grid
| |
A B
where A is the hybrid inverter with PV with CT clamp at a, and B the AC-coupled inverter with CT at b. A is unaware of B, and sees everything to the right of 'a' as "the grid". (Could potentially add an EV between 'b' and the meter, so that neither inverter attempts to power it.)
A is the primary and (by default) doing all the work. B just catches anything A cannot do,
Cross-charging is eliminated because ECO mode on A means it doesn't want any current passing through 'a'. Even if B was to respond more quickly than A to an increase in consumption, A would choose to (also) cover it, then B will back off (because it will become export to the grid across 'b'.)
In the case that A's battery fills up and A is forced to "export" the excess across 'a'. B will charge from it. And if ever A cannot fully cover consuption (because battery has run out, or it exceeds its max or configured output), B will step in to make up the difference (since it doesn't want anything flowing across 'b')
If you want to try to utilise them a bit more equally, you might choose to reduce A's max discharge rate, to say 1kW. Then for any consumption exceeding 1kW, B would chip in. However, would need to keep an eye on B's SoC - if it was to run out, A would not automatically increase output to cover the shortfall. So that's something you could tweak by API every hour or so. And if consumption exceeded what B could provide, A again wouldn't automatically increase its output.
But as has been noted, more efficient to charge A's battery from solar than to send it to B, so you'd want to keep that asymmetry,
This is why I was asking about how you planned to use B outside of the heating season - it doesn't really have much to do. Though I guess on Flux it's probably still better to store excess solar until peak time than to let it out earlier.
Neither inverter can provide accurate stats: only B knows true import/export, but only A knows true generation/consumption.