Another of the GREAT threads that get posted here.
I'm in elementary school comparatively speaking, as far as studying effects of the various forces in play when The Big Light hits the bullet base. On my gas check shanks, I have not strayed far from the Lyman form, that "fingernail's breadth" of clearance between the gas check lip and rear-most drive band. In my GC bullets, this has worked as well as can be expected. I don't ask a lot of my rifles with the castings--1800-1900 FPS only gets exceeded about 10% of the time.
I only have one bullet design with a larger gap, that being the Lyman #454490--a gas-checked 45 Colt SWC. I can't say that it shoots any worse or better than the #454424 or #454190 in revolvers with decent dimensions. Interestingly, "490" was the sole bullet that shot well from The Built Backwards Bisley with its .448" throats and .452" grooves. After the throats were improved, "490" shot about as well as it did previously. It still does, though I don't shoot many of them any more. In cast loads running 1000-1100 FPS, a gas check is rather superfluous. What does this "prove"? I have no idea, that is for the cats in the cubicle farms wearing propeller hats to figure out. Prior to the throat honing, I was happy that at least one cast bullet design shot worth a flop. Even jax were poetic downrange, and not a couplet in rhyme or a hint of iambic pentameter either. Aggravating stuff.