Just a little more dissection of the 22.95 solve, this time LSE.

 I apparently could not wait to do more typing about the 22.95 Minh Thai solve.  Scroll down to the previous post if you're not caught up here.

LSE refers to Last Six Edges. Four of them are in one of the slice layers, and then you have one edge each on the two other layers. Technically, this term didn't exist until the Roux method was proposed, but that's exactly where you find yourself in Minh Thai's method once you have completed three edges each on two opposite sides.

If you look at the last part of the reconstruction from last time, you have this:

u R' E' R E2 R E R' // LSE 
R2 E E' r2 E M2 E' // centers

If you would like to skip ahead to this part of the solve, perform this on a solved cube: 

z’ D2 U2 F' D2 F' U2 F' L2 F' R2 F' R2 F' L' R' U' L' R F D' from standard solved white top/green front. 

So, looking at the moves that are labeled as "LSE", but from the standpoint of Minh's solution guide, I see that breaks down as follows:

u // Aligns upper and lower layers

R' E' R // Puts upper edge in lower target position From Stage III, part 2 (pg 48) "If both (edges) are somewhere in the horizontal middle layer, it will be easy to flip either one of them into the other one's target position."

E2 // puts lower edge in position to be matched with upper Setup move to do Case #2

R E R' // inserts both edges similar to stage III, part 2 case 2.

At this point, all we have is a Dot Case, and there is no edge orientation to do. So, this is a big skip at this point. Perhaps that's why there's the (E - E' ) in the centers because he's excited, or maybe he's just checking to make sure one of the edges in the back isn't flipped first before he fixes the centers. If he had edge orientation to complete, that would have been between 9 and 14 more moves, at least half of which would have been slice moves. Unlike modern cubes, slices can be quite an effort on 80's cubes. So, without that, you get to skip straight to edge permutation, and while it's the worst one, it's not that bad.

So again, we learned that this solve had relatively simple components, a little luck, and was not a  byproduct of a large available algorithm set.