It's probably half and half - I constantly, every day work with colour swatches so I'm used to telling apart minute differences, but it also increases the difficulty the worse your monitor is calibrated, or generally the cheaper the monitor is, or if the VGA ports are damaged. The results can't actually be faked because even though most monitors are blue-heavy or the calibration might annoy some hues, no two colours will cross over eachother or appear the same (but like I said, they can increase the difficulty of the test).