This article uses a bad definition of deck randomization. Instead of basing on entropy/statistics, it asks whether any "two cards have remained in the same relative positions", this definition's utility is not clear (and it is unrelated to randomness / the measure of success of subsequent guessing of cards). Also it states that a dealing method A, B, C, C, B, A, ... gives "improved randomization" compared to the dealing method A, B, C, A, B, C, ... (which is just false since both are deterministic permutations).
> A riffle shuffle (pictured above) requires seven shuffles to randomize a 52-card deck
More generally, it takes a number of shuffles equal to "approximately the square root of the number of cards," per a documentary i watched on the topic several years ago.