R3T

R3T(bad name) is the closest I could get to solving similarly to a square1
Or you could just finish F2L in a normal way.

Contents
Steps(R3TFB)
1)First block+Eslice  Solve a 1x2x3 block anywhere on the cube.
Solve the Eslice by positioning the last 2 Eslice edges.
You can also solve EO first then the Eslice.
2)EO+CO  you solve edgeorientation and cornerorientation.(not necessarily in that order)
 Because the Eslice is already solved, there can only be 0,2,4 or 6 misoriented edges.
4 misoriented edges can be solved as an arrow (like in LSE)
e.g. M'U'M
2 misoriented edges can be solved using an OLL alg.(you can influence
corner orientation in this step.)
 CO can be solved using 12 OCLL algs  7 algs.
(orient the corners stuck in the D layer by doing an R2 first)
This can also be done by inserting the last Eslice edge with TSLE.(Assuming the other D layer corner is oriented l)
(or another subset)
3)Spam F2L  F2L is finished with mainly R2,U,r,M moves.(or L2,U,L,M if it's on the left etc.) Basically pair the last 2 corners with the last cross edge to make the 1x1x3 that goes in the first layer.
4)PLL  Permute the pieces of the last layer  21 algs.
Pros
 FB can be planned in inspection and the transition between steps is quick.
 The remainder of F2L can be solved only using R2,U,r and M moves,spam TPS.
 Alg count is significantly lesser than Full CFOP as you use only OCLL's to orient corners.
 EO can be recognized quickly(maximum bad edges  6)
Cons
 FB is hard to optimize.
 You have to do 2 OCLL's half of the time to orient the corners
e.g. Sune,R2,Sune
 Recognizing EO in the middle of the solve(even though the maximum amount of misoriented edges is 6) will take some time.
R3TEO edge is a variant which reduces the cube into the Spam F2L step by solving EO edge first.
Basically doing FB plus EO differently.

Steps(R3TEo edge)
1)EO edge Solve EO and 2 Eslice edges, they should be adjacent to each other.
e.g.FL and BL(or FR and BR) edges.
2)Pseudo 1x2x3+EsliceUsing the empty faces, solve a 1x1x3 block in the D layer Making a pseudo 1x2x3 block.
(It doesn't have to match the two Eslice edges)
After that, position the 2 remaining Eedges to complete the Eslice.
3)COCorner orientation(CO) can be solved using 12 OCLL algs  7 algs. (orient the corners stuck in the D layer by doing an R2 first)
This can also be done by inserting the last edge with TSLE.(assuming the other D layer corner is oriented)
(or another subset)
4)Spam F2L F2L is finished with mainly R2,U,r,M moves.(or L2,U,r,M if it's on the left)
Basically pair the last 2 corners with the last cross edge to make the 1x1x3 that goes in the first layer,
and solve the DF edges DB before, while or after completing second block.
(You can solve the DF and DB at any time, even before solving the pseudo 1x2x3)
Another way is to insert the 2 corners into the first layer without the DR edge, PBL with Square1 algs(a few a modified to fix "parity"), then insert the DR edge.
5)PLL Permute the pieces of the last layer  21 algs.
Pros
 EO edge can be planned in inspection and the transition between steps is fairly quick.
 The remainder of F2L can be solved only using R2, U and M moves, just spam TPS.
 Alg count is significantly lesser than Full CFOP as you use only OCLL's to orient corners.
Cons
 EO edge is hard to optimize.
 You have to do 2 OCLL's half of the time to orient the corners
e.g. Sune,R2,Sune
 Easy to get lost between steps as the steps aren't straight forward.
Tips
 In the spam F2L step, instead of only using R2,U to comple F2L,use also r2,U and M,U
R2 U R2
r2 U r2 U'r2
M2 U M2
This will increase the efficiency of this step.
Variants
R3TZZ:EOEdge,pseudo 1x2x3,EOLine,F2L,LL
R3TRoux:EOEdge,pseudo 1x2x3,SB,CMLL,LSE
R3TEO edge:You reduce the cube into a domino state by starting with EO edge.
R3T4c:Solve the UL and UR edges in the DF and DB positions before or while finishing F2L, and end with 4c after permuting the corners.
R3TLSE:You don't finish the entire F2L, instead, you solve the second block with R2,U and M, then permute the corners while influencing the UL and Ur edges and end with the rest of LSE.
(another way is to solve CP sometime before and solve CO without doing F moves)
Overview
I don't think it's comparable with bigger methods like CFOP, ROUX, ZZ and Petrus, but it's definitely fun to use
You solve F2L in a ridiculous fashion  solving the Eslice(which many people don't like) and doing CO before finishing F2L.