This is just a brief story about my experience trying more advanced methods for edges in blindfolded solving. I normally use M2 for edges and Orozco for corners. I decided I wanted to try out Orozco and Eka for edges, to see if I prefer one of these over using M2.
I should note before I continue, that I do use a lot of special cases when doing M2. Rarely will I do (setup target 1) M2 (undo setup), (setup target 2) M2 (undo setup). Instead, I tend to solve two pieces at a time. A lot of my cases involve setting up to U M’ U2 M U with R/L moves, or setting up to a commutator with E or S-slice interchange like R’ F’ R S R’ F R S’.
Anyway, after trying Orozco and Eka, I decided that I liked Orozco better (likely because I’m so used to the logic after using it for corners). I also had trouble remembering my setup moves for Eka, and although this would have gotten better with practice, I didn’t like the idea of setting up to a suboptimal commutator. So I continued trying it Orozco in solves, until I realized that I am much slower using Orozco rather than M2.
At first I had thought this difference was due to a lack of experience with Orozco commutators. Surely with more practice, I thought I would get faster. However, I realized that the problem was that Orozco is still solving one piece at a time, whereas with M2 (or rather, my variation on it), I was solving two pieces at a time for all but the easiest standard cases (mainly when one of my targets is A/UB).
All this to say that just because a method is more advanced, does not make it faster. Also, it is never too early to start incorporating commutators in your solves, especially if it’s for easier cases like the ones I mentioned above. I think familiarity with the letter scheme is much more important for becoming a fast blind solver.