Smoke and mirrors on Guantanamo recidivism

Remember all those fevered claims that 20% of the men released from Guantanamo had "returned to the battlefield"?  Today, Marc Ambinder's report on politically motivated Republican attempts to keep prisoners in Guantanamo contains this tidbit:

The letter doesn't refer to what the intelligence community has concluded about recidivism, but a senior U.S. official tells me that the Defense Intelligence Agency has concluded that just 11 of the entire class of 525 former detainees have returned to the battlefield in some capacity and are still at large.

That contradicts Ambinder's own fairly recent report on the same subject -- unless some 80% of alleged recidivists (discounting 'suspected' ones) have been recaptured or killed:

The recidivism rate of the 558 official Guantanamo detainees is hotly debated. The Defense Intelligence Agency confirms that about 10 percent returned to battle or terrorism; an additional 11.2 percent of the released detainees are suspected of having done so.

A small fraction of the number, however, remains at large -- so most of those confirmed to have returned to terrorism have been recpatured, [sic] are imprisoned, or have been killed.

Mark and Joshua Denbeaux et al, attorneys to Guantanamo detainees,  have demonstrated in exhaustive detail that Dept. of Defense estimates of Guantanamo recidivism are vague, undocumented,  mutually contradictory  -- in short, completely unreliable. Which is not to say that the figure floated by Ambinder's latest anonymous source has any particular claim to credibility.