It could be measurement error combined with a little wishful thinking, or it could be that there's some undiscovered physics there and we're tip-toeing around the edge of the parameter space where it manifests. If that's the case, all these "X effect" systems could be working according to the same principle.

There are other cases in science (e.g. early transistor efforts and superconductors) where tinkerers and engineers hit on effects that were not understood and were ignored for a long time until we had some kind of theory that explained them and much more importantly told us how we might optimize for the effect.

The McCullough prediction that a dielectric should increase the effect seems easy to test.