The SC has not ruled as you desribe. The SC has no rule that requires a murder IOT sustain a death penalty, and it certainly does not have a track record of overturning death sentences just because a crime isn't particulalry gruesome.
Gregg v. Georgia (1976) rejected the death penalty schemes of two states (NC and LA) because those statues didn't allow sufficient discretion in applying the death penalty. (Basically, it came too close to a mandatory death sentence for a specified crime, something the court rejected.) In rejecting the "cruel and unusual" argument, the court said that the death penalty isn't always cruel and unusual if it isn't disproportionate to the crime. (In other words, it
is cruel and unusual unless the crime is proportionately heinous and unusual.)
In 1977 Coker v. Georgia found that the death penalty was grossly disproportionate for the crime of rape of an adult.
In 1984, though, Pulley v. Harris rejected the effect of Gregg v Georgia that required a comparative proportionality review, "[a]ssuming that there could be a capital sentencing system so lacking in other checks on arbitrariness that it would not pass constitutional muster without comparative proportionality review". In other words, if there was a proportionality problem, it could still be reviewed, but the review was no longer mandatory in all cases.
In 2008, Kennedy v. Louisiana rejected the death penalty (as being disproportionate) for the rape of a child where the victim did not die. It ruled that in crimes against individuals, the death penalty cannot be used unless there was a homicide or attempted homicide.
So, with the "aiding the enemy" charge, the death penalty can still be used without an attempted homicide because it's not a crime against an individual, but there is a clear tendency in the Supreme Court to reject disproportionate application of the death penalty.