Sure but I don't immediately see how that equates to a proof of its ability to produce every combination of, say, 52 consecutive numbers... Just that it will pr

Sure but I don't immediately see how that equates to a proof of its ability to produce every combination of, say, 52 consecutive numbers... Just that it will produce every possible individual number. In PRNG-parlance it shows 1-dimensional equidistribution but not 52-dimensional (or generally (k, w)-dimensional) equidistribution. Or is there another pigeonhole step I'm missing?