Can we filter results of mapping tools by percent-identity and aligned-percent? And, is conducting such filtering appropriate?

I thought that mapping tools like minimap2, bowtie2 or bwa-mem didn't have such options, but I found that COVERM, that conduct both mapping and abundance calculation, has such options (COVERM seems to use minimap2 or bwa-mem). So can we filter SAM or BAM by percent-identity or aligned-percent? Also, since I don't see many people doing that, is it inappropriate to filter by such BLAST like criteria for mapping results?

I am trying to map metagenome data to references and am wondering if this kind of filtering is a good way to determine the sensitivity.

Thanks.