The CF 5.0 specification combines the old 28-bit addressing with 48-bit addressing, which not only increases the capacity limit, but also allows data to be moved in larger and fewer chunks to theoretically achieve better throughput. In addition, the spec implements commands that allow the host device, such as a camera, to better manage unneeded data and deleted file spaces, which the CFA claims will fix a CF card's tendency for performance to degrade over usage. (Will this affect the efficacy of deleted/corrupted file recovery software, I wonder?) There's also some changes to its controller standardization to streamline future interactions with forthcoming standards.
But what you'll likely hear and see most about are the "Video Performance Guarantee Profiles." These will allow the card's controller to communicate to the host device its best- and worst-case performance parameters and optimal transfer characteristics. The consumer-facing aspect of these are the buckets that marketing folks will be able to pour the different cards' performance into: HD (10 megabytes per second sustained transfer), Pro (20MBps) and 4K (75MBps). The HD bucket is equivalent to SD Class 10. Now, is it just me, or wouldn't it be OK if the manufacturers simply said, "You need performance of at least X MBps to shoot video," since the card manufacturers are already reporting these throughputs? I find these classifications tend to confuse rather than clarify--especially in this case, where the more common rates are 15MBps, 30MBps, and 45MBps.
Let's hope clarity wins by the time we start to see the first cards. I'm not sure when that will be, as yet.