It seems to me that Scheuer is really stretching to make his supposed point here.

Several of the prisoner releases that he cites are overwhelmingly criminal, not political, prisoners. In these cases, most were nearing the end of their sentences, and their release (to mark religious and national holidays) has come to be as much an expected part of the judicial and incarceration system as is parole in the US.

Second, Arab regimes are--precisely because of blowback from the Afghan experience--fully aware of the dangers of militants traveling out of country to receive "work experience" in a foreign insurgency. It stretches credulity to expect that most would today see this as an effective strategy for limiting domestic national security challenges given their knowledge that most of those chickens come home to roost.

Third, some of those prisoner releases have to be seen in the context of successful deradicalization efforts by Arab governments. In Algeria, for example, national reconciliation efforts have reduced the militant Islamist movement to a tiny, tiny fraction of what it was during the height of the post-1991 civil war (a conflict that claimed well over 100,000 lives). Egypt and Saudi Arabia have also experienced some substantial deradicalization successes, in part because of their manipulation of selective inducements such as early release.

Certainly there are recent cases of released prisoners going on to cause mayhem elsewhere (Shakir al-'Absi of Fateh al-Islam comes to mind), and there is particular grounds for concern regarding Yemen's detention-and-release (or escape) policies.

However, I suspect that a look at the background of captured foreign jihadists in Iraq in particular would show that very, very few of them had been imprisoned Islamists given early release in their native countries. On the contrary--and rather more alarmingly--a significant proportion were rather latent Islamists mobilized into action by US intervention in Iraq.