Yes. Though, it isn't the width per se, but the generally corresponding lower tire pressure. The lower pressure allows the tire to deform around the bump, rather than the bump lifting the entire bike and rider into the air. The force that goes into raising the elevation of the bike/rider is essentially force that is deducted from the forward kinetic energy of the bike/rider.

I think the important takeaway from all of this is that the tires are close enough to each other that it is possible for the speed difference to be obscured by the margin of error in an experiment. Granted, the margin of error of the OP is large, but other tests (bikerollingresistance) have shown the tires to be quite close as well. The Hypers very well might be objectively faster on crappy roads. (I've had 40mm hypers and 28mm GP4000s for what it's worth).

Having had both tires...I generally agree with the assessment of the OP. I feel like the GPs have a slight edge on good roads, but it really is so close as to say any difference could easily be attributed to my imagination.

On bad roads, I'd say the Hypers are definitely faster. They roll faster, and you do feel more confident going over broken pavement, pedaling over stuff you would unweight yourself on with the GPs.