If you take this idea to its logical conclusion, you end up with Predictive LR(k) (or PLR(k)) grammars:

https://link.springer.com/article/10.1007/BF01934444

These are essentially the LR(k) grammars that become LL(k) with a combination of a left-corner transform (to eliminate left recursion) and factorization of common prefixes.