Might be they estimated based how quickly bacteria eats pure metal, but it could be covered with something that protects it, dust, rust, another bacteria, some chemistry that is synthesized there over decades.
My gut tells me the biggest source of error is they probably estimated it in a clean-room environment for the bacteria.
Bacteria in the wild are perpetually in a war for survival with other bacteria (and larger organisms) that see them as an incredibly convenient source of aggregated biomass and carbohydrates. I'd expect the rate at which they chew through metal in the wild is attenuated by predators on them increasing predation as their numbers grow.
