I'm in the process of redesigning the MGLS database. (The test index is
here, if you care to see.)
Im going to add some more information and note that MGLS seems to be excellent for OH and big cubes. 2-gen is fast for OH, and if you've seen Frank turning a 5x5x5, you'll note that a firm grip allows for best speed with 2-gen (alignment is less of an issue) -for a 7x7x7, 2-gen is even better.
I'm not sure if there's any significant improvement over Fridrich for 3x3x3, but we'll see. I'm working on it.
Anyhow, I adopted MGLS to see what it could do, and the apparent potential is encouraging. Imagine what happens if a few cubers use it and each is on the hunt for fast algs.
The thing is, ELS is comparable to last slot and CLS to OLL.
Theoretically, CLS has a lower move-count than OLL. However, you an choose longer algs for better speed, efficiency, confort, smoothness, etc.
So, either you use fewer moves or faster algs. (And seriously, the number of algs is not that bad compared to Fridrich, especially if you take into account mirrors.)
Now, here's the great part:
What if you had the choice of replacing any of your OLL algorithms with a 2-gen alg? Your worst cases will magically turn into a 14-move 2-gen (or maybe a 10-move 3-gen alg). MGLS is like that. At first, I shied away from unnecessary 2-gen, but I'm switching over. It's great.
For OH, CLS has a lot of nice 2-gen cases. Most of my OH is bad (that's mostly just cause I haven't refined and practiced), and I tend tend to make devastating mistakes, but there are enough nice cases that in my 30sec averages there is always a really low 20 (and you can make your ELS easier through care on F2L).
My official 21.71 and 20.93 OH solves were that way. The 20.93 was also a PLL skip, but CLS was ridiculously fast; I had replaced a very bad alg (-23) with a much nicer one. The rest of my solves at Berkeley were bad, though, due to PLL, some leftover bad CLS algs (I still haven't relearned some from my original set), and other issues...
(By the way, I've done two 29.xx OH averages...)
.
