Hi @Megan Justice 👋
Some fast answers:
1. Currently not no
2. Not by taxprofiler itself, from the top of my head some of the tools may calculate percent read count abundance
3. Technically yes you could, but like you said they should really be normalised. I don't remember if nf-core/differentialabundance's deseq2 step actually does normalisation or not
But as a general thing, taxi metagenomic/microbiome profile normalisation remains active area of debate. People still argue what is best out of rarefaction, log-ratio-transform, ancom, deqseq2 like methods so you have look into each one and decide which is best for your context. This is why we have not added our own steps to taxprofiler itself as we don't want to misguide people