This is an irrelevant distinction for any case where you aren’t worried about someone reverse engineering the algorithm and seed by logging output. Any half decent PRNG’s output will be statistically indistinguishable from true randomness.