I posted a couple threads like this ~1 year ago, I'm pretty sure i did it because I was worried ZZ was holding me back.
Since then I've got sub 15 wth the bug 3 and I can comfortably sayt hat you adapt to your method to the point that he one you use doesn't matter too much. I think that even...
On the other hand, I cant stand magnetic cubes for oh, i find them too uncontrollable, it's really just personal preference but until you get sub 20 oh it dosen't really matter
in that case let me (a pro) help you!
this step is fine nothing to say about it
this is the route of alot of your problems for the method (in my pro opinion), in the two best cases you will have df or dr unsolved, however you have solved 3 pairs at this point making it awkward to do the next...
so why not use a petrus/roux/zz/pcms/m-cell/freefop style f2l
ollep probably isnt the best idea, as if you use oll you only need to see 2 sides but the increase in recognition time for identifying the 3rd side will take too long to be wroth it as cpll isnt exactly amazing