As some of you may remember, I proposed a Petrus based CP-block method a while back. The problem with it was that it made Petrus even more inefficient than it already was just in order to have a 1LLL.
I never dropped the idea of CP-blocks, however...
...and today I bring you my NEW AND...