Storage of size of block instead of bucket

I just wanted to know if there is any specific reason in the choice of storing the size of the requested malloc() block in the header, instead of directly storing the bucket this buddy would fall into (so we don't have to recompute it with a loop).

Thanks!