diatoms are 100% independent of cycling completion, they're incidental early hitchhikers in the system. tanks that use live rock transfer setup might not even get them
bare bottom systems rarely get them at all, its sandbed relations and silicate that many think hold the link.
one part that seems to be pretty recurring timing is that diatoms show up about a month after water addition, which is roughly the time bacteria have deposited and reproduced on surfaces, so people have associated diatoms with cycling completion based on that rough estimate timing, but it wont always work that way. they may show up earlier, or later, than full ammonia oxidation does in a given system's surface area we are cycling.
the definitive test for cycling is an ammonia oxidation test, ran after a large water change in a given system (so that bac added in suspension aren't giving a false reading, we want surface area readings) to test for the ability to oxidize 1-3 ppm within 24 hours, that's the close or not of a given cycle.
*its possible to instantly add fish to a tank though with todays additives, such that no free ammonia exists on day one even with fish...although this is zero ammonia on a given bioload, the system cannot endure a complete water change and resume oxidation, as only the additives in the water were doing the work and would have to be added back so we don't really call that cycled, it takes minimum about 2 weeks to get a set of rocks truly cycled and that's if you precisely drive the cycle with exacting ammonia and bottle bac dosings.
