So far the correlation between personnel proximity and tank depletion is just that: a correlation. All attempts at establishing causality have been met with fuzzy recollections, outright denials, and conflicting reports.

More investigations are warranted.

The earliest HDDs were not sealed and required regular maintenance to clean the heads manually; you were supposed to use ethanol or isopropanol for that purpose. There was a story from late Soviet times about this - since they lagged quite badly when it came to computing, even in late 80s there was still plenty of hardware around like that. And for every such hard drive, there would be accounting to track the amount of ethanol issued, so the engineers had to make do with what they could skim after the cleaning.

But then around that time some better stuff began to trickle in - instead of 3-5 Mb HDDs, you could get a whopping 30 Mb! Those were sealed and required no cleaning... but the legend goes that, in one such place, it occurred to the engineers that accounting doesn't know that. So, instead, they submitted a request to procure more ethanol for cleaning, on the basis that a 30 Mb hard drive would obviously need 6 times as much for servicing as a 5 Mb one does. It was approved, and the arrangement lasted until the head of the department caught someone drunk and traced the source.

(Another version of this legend has a slight variation in that the engineers submitted a request for more than 6x, deliberately making a mistake in math to make it look like a clumsy attempt to get more. Accounting promptly "caught" it and gave them a stern lecture about how they're not getting even one drop over what the proper number required... i.e. 6x.)

If you liked that story, you'll love https://www.youtube.com/watch?v=bKoHMXggEHU