IMS withdrew the QTI 2.1 spec, despite some existing implementation and information going back to 2006 that is was imminently ready for use. The official IMS wording seems to be toned down a bit, but there is also a clip from early wording on Rowin Young's blog. Other early opinions ended up on list servers like this. I've fired up a Google alert on this issue and will be tracking it. Meantime, one of my co-workers at Questionmark and a key contributor to QTI 1.x, John Kleeman, has penned, Why QTI Really Matters.
Check it out and watch for updates here. I expect a few more QTI experts and pundits from other specification bodies to have some observations soon. However, an unnamed organization with a litigious nature may be unwittingly suppressing discussion.