The difference is that this method that I'm describing does something different with the final pair, and uses a wildly different ZB subset for the LL.
Also, how long has that ZZ thing been around? I came up with this (we've been calling it M3) almost two years ago.