-----------------------------------------
Step 2: Create DFR/FR pair then orient U-layer corners while inserting
Step 3: Permute U-layer corners while solving a D edge
Step 4: E15 (created by Kenneth Gustavsson)
Problem: Not orienting edges after Step 1 causes more memorization for the rest of the steps and doesn't seem to change the speed. You still have to orient the edges during E15 and it has a higher move count than doing it as the second step.
-----------------------------------------
2. Orient
3. DF/DFR/FR (fun step)
4. Solve corners
5. Edge permute
Problem: Seems faster than the regular method but it doesn't feel unique, feels more like "Roux" + Heise.
-----------------------------------------
2. DF/DFR/FR
3. Solve corners
4. E15
Problem: Same as the above, taking out the orientation step doesn't seem to change the speed.
-----------------------------------------
2. Orient
3. DF or DR and separate edges
4. Orient corners
5. Permute rest (may be too many cases)
Problem: Unique enough but Step 3 may be a lot of cases and Step 5 looks to be PLL times 5.
-----------------------------------------
Expansion:
2. Orient
3. If pair is solved, use COLL.
4. Permute remaining edges
-----------------------------------------
Expansion:
2. Orient
3. If pair is solved, orient corners and separate edges in one sequence
4. Permute rest (probably 42 cases, either DF and DR are solved or they are switched)
-----------------------------------------
And many others that didn't seem as good to me.