You're still having to do all the parts to LL: co, cp, eo, ep, but you do them all separately in the wrong parts of the solve. You're just moving the parts of LL into F2L.
Would you consider Roux a LL skip method? It does exactly this but with fewer algs and better flow overall.