Why does setting lfcThreshold > 0 in lfcShrink() of DESeq2 report s-values instead of p-values when using apeglm shrinkage?

Currently using Bioconductor 3.10, DESeq2 1.26.0

The p-values are from a null hypothesis test produced by results, but lfcShrink can also produce s-values, and it can test against an LFC threshold. It is just confusing If you want threshold test p-values and posterior estimates of LFC together in the same table, I'd just make two tables (one with results and one with lfcShrink) and then add the relevant column from one to the other.