I’ve been following James Sulak's Blog. He has some pretty impressive detailed discussions about using XProc. XProc is an XML pipeline processing language, specifically designed to provide instructions for processing XML content. The Recommendation specifies many different kinds of “steps” that can be assembled in virtually any order to control the sequencing and output from one step to another.
Right now, DITA’s reference implementation, the DITA Open Toolkit (DITA OT) uses Apache Ant and custom tasks to process DITA XML content. One of the principle limitations with the DITA OT is its reliance on XSLT 1.0 and extensions (particularly the Idiom FO Plugin) to handle the rendering.
With XProc-enabled tools like Calabash, it seems like DITA could easily processed using XProc, along with an upgrade of the stylesheets to 2.0.