Optimizing and improving the speed of a recursive algorithm that compares strings

Question

The following algorithm works as such: Given two strings, return true if the strings are equal to each other regardless of their vowels.

"water" and "wtr&q