The way I understood this article, the problem Fibonacci hashing seems to solve is that it turns a hashing strategy that would require a prime modulo into something that can use a power of two modulo.

I think there are some hashing functions around that are already designed to solve that problem at "step 1".

So the question just boils down to which is faster