Reproducing the image in this way is essentially a manual low-pass filter (although with little control over the parameters), so it's certainly one valid data p

Reproducing the image in this way is essentially a manual low-pass filter (although with little control over the parameters), so it's certainly one valid data point with which to test the hypothesis.