I was concerned when I read that one too, as I specify and recommend SLC DeviceNet scanners by the truckload.
But I've never written a program that wrote Explicit Messages prior to copying Output data to the M0 table. I usually run a bootup timer and check the Device Failure Table before I send explicit messages. Meanwhile, the logic that copies data to the M0 file has run hundreds of times.
I can see where you might have periodic or on-change COP instructions to the M0 file and this could come into play. Bugs that cause outputs to not be set to the correct state are always considered high-priority service alerts.
It's a good upgrade but not a cause for panic.