Mutil is a set of standard utilities that have been parallelized to maximize performance on modern file systems. These currently include multi-threaded drop-in replacements for cp and md5sum from GNU coreutils, which have achieved 10/30x rates on one/many nodes.
Looks like an excellent solution to my data movement problems. Wish I could get it to compile. It relies on coreutils 7.6 which is quite old and I can't seem to get to compile on any of the systems I've tried thus far. (Centos 6.5, 7.0, Ubuntu 12.04, OmniOS) A little bit of documentation would be helpful.