Benjamin and Emily appear to agree that, as he puts it, "[ t]o the extent the eventual convictions of KSM et al rely on coerced testimony, even indirectly,... the Defense Department should not put them to death ." But should the prospect of execution alone be the only concern? Benjamin does proceed to discuss "clean convictions," implying the answer is "No."
At least since the days of Mapp (1961) and Wong Sun (1963) -- or, for that matter, Bram (1897) -- the question of tainted evidence has arisen 1st and foremost at the guilt/innocence phase. If it's addressed properly there, most likely there'd be no cause for reconsideration-in-mitigation at sentencing. Seems a simple enough premise. Yet it's one away from which the U.S. criminal justice system's seemed to have moved in recent years. Example of this shift: the widely shared notion that it's a victory when a tainted-for-whatever-reason capital sentence is commuted to life.