The purpose of "batteryless" direct grid tie is to feed the grid for the tariff or tax benefit, or to reduce your electricity bill. If that's the goal then it will depend on a lot of factors that I don't know anything about - what country/state/county you live in, for example. None of that is intrinsic to the equipment you already have, so it's just an economic decision. Therefore it can become a somewhat separate project that you can optimize for its own sake without much of an interface with your battery based system (except at the final AC interconnect).
Since you already have a battery-based system (that I envy) you have backup power, if that's a concern, already in the bag, so you don't have to complicate your grid-tie battery system by splicing the DC together.
With a second flexmax you will be optimizing the efficiency from the 3.5kW array anyway, just like the Sunny Boy would do. As much? Only someone who has examined a comparable system or done detailed analysis could know the difference. But I bet it would be close. So again I say your personal economic decision rules this. Only you can make that decision.
Competition is fierce in the grid-tie inverter market, so what you saw in a product review last year may have changed by now. Get the most up-to-date info as possible, and don't forget that there are several other viable competitors out there, though SMA has a good reputation.