The crazy thing about the influenza virus is that it's genome is fragmented into 8 pieces. Now, imagine a single cell is infected with 2 strains of a virus... Say the aforementioned H5N1 virus and and this H1N1 virus have both infected one host. Now there is a very real possibility of rearrangement between the 8 viral RNAs of those two strains. BAM! You now have a very scary virus, and very quickly. Another thing that has been neglected in this conversation is that viral mutation rates (in viruses with RNA genomes anyhow) are MUCH higher than what we normally consider, as RNA-dependent RNA polymerases have zip for proofreading activity. Yep, I'm going to wash my hands 42 times now.

Science doesn't really follow Occam's Razor, particularly in evolutionary biology. Many processes are not parsimonious. Go look up systematics, or look at some phylogenetic trees made with modern methods (e.g. Bayesian inference or maximum likelihood, not maximum parsimony). Occam's Razor will fail you.