How Does a 14% Guantanamo Recidivism Rate Strengthen "the Arguments of Critics Who Have Warned Against the Transfer Or Release of Any More Detainees"?

That quote from the NYT. In rather more detail it looks like this:

An unreleased Pentagon report concludes that about one in seven of the 534 prisoners already transferred abroad from the detention center in Guantánamo Bay, Cuba, has returned to terrorism or militant activity, according to administration officials.

The conclusion could strengthen the arguments of critics who have warned against the transfer or release of any more detainees as part of President Obama’s plan to shut down the prison by January. Past Pentagon reports on Guantánamo recidivism have been met with skepticism from civil liberties groups and criticized for their lack of detail.

Right, well, the first question that comes immediately to mind is:

1. How on the damn Earth does a 14% recidivism rate strengthen arguments against releasing apparently innocent prisoners?

And, um, since--unless someone out there is arguing for releasing apparently guilty ones--that's what we're really talking about here.

The second question that comes to mind is:

2. What are ordinary recidivism rates like?

This question is raised and answered--sort of--later in the story. The answer given is: "as high as" 68%...though unless you're particularly gullible, you probably won't believe that figure. But suppose it were 50%. Heck, suppose it were 30%. If so, that 14% rate still doesn't provide us with a decent argument against releasing the apparently innocent among the Guantanamo detainees. This should be too obvious to need saying, but: the question that should determine whether detainees should be released or not is: are they guilty?

(Note that one thing we seem to be assuming here is that this is actually recidivism we're dealing with. It would be very interesting to know whether any of the people who engaged in terrorist training etc. after their release had been non-terrorists before their imprisonment...)

Perhaps the argument is that these folks are likely to be so dangerous that even such a low recidivism rate is unacceptable. This seems extremely implausible to me, but I'm willing to be convinced about the point.

And finally, of course, one wonders:

3. How does any of this provide any kind of argument for transferring prisoners to normal American prisons?

The answer there, of course, is simply: it doesn't. None at all.

Finally, it's worth noting that the Pentagon got this 1-in-7 figure through a complex method of investigation that civilians know as "making shit up." The real number is apparently much lower:

The Pentagon has provided no way of authenticating its 45 unnamed recidivists, and only a few of the 29 people identified by name can be independently verified as having engaged in terrorism since their release.

Though there's some Times-y hedging immediately thereafter:

Many of the 29 are simply described as associating with terrorists or training with terrorists, with almost no other details provided.

Ah, right--merely training with terrorists! Well nothing wrong with that! [/sarcasm]. See, that's good enough for me. "Simply" training with terrorists is something I'm willing to count as recidivism here. Let's not get overly picky just to make the story a little juicier, shall we not? So I'm willing to accept the '29' figure (which includes those who merely consort with terrorists)...but that gives us a recidivism rate of closer to 1 in 18.4, not 1 in 7.

Maybe I'm just missing something here...