The graph shows what happens when researchers desperately hunt for statistical significance like it's the last coffee in the lab. See those suspicious peaks at exactly z=1.96 (p=0.05) and z=2? That's not nature's joints—that's researchers frantically massaging their data until it coughs up a "significant" result.
This is the statistical equivalent of fishing with dynamite. If results were honest, we'd see a smooth curve. Instead, we get these magical thresholds where suddenly EVERYTHING becomes significant. Thirty years in academia and I've never seen nature organize itself around arbitrary p-value cutoffs!