ECP Method

ECP is a method that was invented for the Method Development Competition January 2021 by team Kübirz. It is designed to use inspection to do an intuitive step, in order to get to algorithmic steps as quickly as possible. Experience with ZZ/CFOP last layers, Roux’s L4E, and ZZ’s Edge Orientation step are highly recommended.
Steps
1. Solve EO and build a face on the D layer.
2. COES, Corner Orientation & Edge Separation, orients the U layer corners while separating E and U layer edges into their respective layers. This step takes 111 algorithms.
3. P3L, or Permuting 3 Layers, is the next step. 1looking this step is not feasible because it would require the memorization of thousands of algorithms. Instead, this is broken up into 24 steps depending on which variant you want to use. The recommended variant is called ECPCPEA. It 3looks P3L with 59 algorithms. The recommended advanced variant is called ECPCPES, which 2looks P3L in 178 algorithms.
ECPCPEA
3a. CPEA, or Corner Permutation & Edge Alignment, is a new algorithm set with 35 algorithms that permutes the U & D layer corners while also pseudo permuting the edges such that only U, H, & Z perms can come up in the next step.
3b. EP, or Edge Permutation, consists of 24 algorithms that permute the edges of the U & D layers.
3c. PES, or Permuting the E Slice, is the final step. This can be done in exactly the same way as in the Roux method.
ECPCPES
3a. CPES, or Corner Permutation & E Slice, would contain approximately 79 algorithms (they have not been generated yet) that solves the corners of the U & D layers while also simultaneously solving the E slice.
3b. EP, or Edge Permutation, permutes the U & D layer edges. There are 99 in this variant, as opposed to 24 in ECPCPEA because Edge Alignment wasn't solved. Also these algs are different because they aren't allowed to destroy the E layer.
Pros
 After the 1st step, pretty much the entire rest of the solve is algorithmic, so there is very little slowing down to do intuitive moves/thinking
 Using inspection to plan intuitive steps is a good use of inspection
 Rotationless (technically)
 COES algorithms are mostly 2gen <RU>, allowing for fast executions
 EO gives the solve better ergonomics overall
 Due to the high amount of algorithmic steps, overall TPS is increased
 Additional algorithms could be generated to avoid AES
Cons
 P3L recognition requires looking at at least 3 sides, which is slower than something like 2sided PLL recognition or CLL recognition
 No pieces are directly solved until the last step, which increases blindspots
 Permuting the E slice either requires either E moves, or a rotation
 You must be familiar with PLL recognition to recognize CPEA, but PLL isn’t even used in this method (Except for in the ECPPLLP variant)
 More algorithms than CFOP and Roux combined
Improvements
 New COES algorithms could be made that ignore the permutation of the D layer, allowing for algorithms with a lower movecount overall
 Additional COES algorithms could be learned to avoid having to Adjust E Slice (AES) before algorithms.