RNA-seq experiments permit to obtain the complete CDS of a large number of expressed genes, thus allowing large-scale phylogenomic studies. Generally, the use of large sequence datasets in phylogenetic analyses provides more trustworthy and accurate results compared to estimates based on a low number of sequences. RNA-seq comes particularly handy when complex evolutive nodes have to be resolved (take a look at http://www.nature.com/nature/journal/v480/n7377/full/nature10526.html for an example)

Interesting topic. Mapping the reads onto the other species can give some answers about the homology and phylogenetic distance, as Marco says. You can build a table of homologs and differences between species - probably quite easily.

It is not that much explored, but perfectly publishable topic - so good luck !

There was also a paper from the Carnici group about the evolution of the expression of the Toll receptors. Maybe this will help: http://www.ncbi.nlm.nih.gov/pubmed/22451944