Will that prevent microstuttering? sounds good btw.
If they both operate on the same frame, then yes

But, as they say up norf... "you don't get owt for nowt". Having multiple GPUs operate on the same frame is generally not as efficient as having each GPU operate on a different frame. There is an additional overhead associated with the analysis of the frame and the distribution of the workload, and an associated increase in communication overheads. Given this, I'm expecting that it will scale less efficiently than existing crossfire or SLI methods.
That said, unless the difference is massive, in my eyes it will be well worth it. I've never been a fan of AFR.