How should we compare different genomic estimates of the strength of inbreeding depression?