I'm quite certain that molecular/atomic SIZE makes a huge difference. For example, if you were to look at standardized gasket testing, you would see that they use helium (second smallest atom). The helium leak rates will be more than, say, water or even air. However, helium leak detection requires a very sensitive "sniffer" whereas water is pretty easy to detect. You will also notice that any flanges in hydrogen service will not have insulation boxes around them because the hydrogen leaking through the gasket would accumulate in the insulation boxes and could lead to an explosion.
If it's a sensitive leak test that you're after, then a low pressure air test with soapy water (to detect bubbles from the air) is probably your best bet.