I believe he's using checking springs that are too soft for flow bench testing. That would explain the pattern I've seen where all his posted flow tests have abnormally high low-lift intake #s.