That seems very believable in a general sense, and I have vague memories of other studies that suggest similar things.

The remaining question for me is whether and how EffectCheck's results have been compared against actual measurements of these subliminal effects.