I thought the IQ8 marketing material specifically says it has a much faster microcontroller/lower latency/finer grain control loop controlling the inverter output vs IQ7, to handle load matching.
This is also used as an explanation for why their system needs a particular minimum ratio of IQ8 to IQ7 microinverters in some configs.
I mean, sure, take with a grain of salt, I guess.
iQ8 is certainly faster but not that relevant to the AC coupling issues being discussed. The main AC coupling issue in an island grid with iQ8's is the lack of full initial sync and power ramp up with the grid-forming inverter (GFI). It happens well before there is excess power beyond the load consumption or the charging power limit of the GFI. The likely cause is due to the iQ8's grid disturbance injections (anti-islanding, sunlight only backup testing, etc.). My speculation is about when Enphase home batteries (e.g. iQ10) are paired with iQ8 micros where proprietary comms are likely used to disable iQ8's grid disturbance injections in the island grid and curtail power without frequency shifting. Such scheme would allow for much more stable island grid and no annoying side effects of frequency shifting (e.g. flickers, clock drift, etc.). BTW, Tesla Powerwall works in a similar manner when paired with Tesla PV inverters. I suspect PV inverter makers are simply leveraging the out of band power control functions they developed to comply with UL1741SB. It's just too bad that the comms interfaces are not made accessible to owners. If this trend continues then AC coupling with third-party GFI (e.g. Schneider, etc.) will likely get worse.