Ployd, everything you mention sounds reasonable. The same things we saw in our etherless experiment thread. The only thing missing is the likely small loss of power by running a higher percentage of ether.