Currently, the only way to use altHypothesis in DESeq2 is through results. However, this doesn't allow to use altHypothesis in shrunken LFC values. I want to isolate genes with low LFC and I'm using altHypothesis="lessAbs" for that, but, as my LFC values were not shrunken, I think I may be underestimating the fraction of genes with low LFC.

You don't want to use shrunken LFC for when your alternative hypothesis is that the genes are less than theta in absolute value (and the software won't allow this), see the DESeq2 paper for discussion.