wanted to say thanks for the tips, Paul. I was able to narrow down some things about my motors stuttering , inspired by you 'asking the right questions' 
basically I tuned the PID controls on the onboard microcontrollers and got acceptable performance/smoothness. The issue didnt seem to be the serial throughput after all. Your suggestion inspired me in this direction. Cheers.
As for the 'P' in my data stream that was a placeholder for a future feature. I wanted to 'stress test ' the amount of serial I could send. My current thought is that serial is probably not a bottleneck as much as I thought.
I will be going back to my dynamixels soon enough anyhow, as they have the ability to send high res position info. Will definitely be more serial 'fun' ahead.