Reading between the lines here, the situation seems to be a little like the softmodem situation, which was also a major pain for Linux support. The abstraction previously given in hardware has been split into software/hardware to save on hardware costs, but now they refuse to open source the software.
As for the provided user-space software, it's not clear why this is happening. As long as kernel modules keep to the allowed interfaces, (and not the GPL-only internal interfaces) they can be closed source. Is it a case of too much caution from the vendors (after all -- NVidia have closed source drivers)? Is this supposed to be more forward compatible? Is it possible to have a higher performance (zero copy?) user-space exposed abstraction for this kind of video processing?