The Manhattan Project, as I understand it, employed 100,000 people (when 130-140 million people lived in the U.S.) and likely all hundred thousand were bound to keep the Project's work secret, probably several foreign governments were skillfully and resourcefully trying to uncover it, at least one did uncover some of it, and at least two had similar projects of their own, and the President of the United States knew of it. The President was sick and was expected to die in office. But the Vice President of the United States did not know about the purpose of the project until after he became President himself. To this day, the scientific validity of the secret project is not much denied and, in the U.S., neither is the military validity of keeping it secret, given the war then underway. Yet the information did not leak to Harry Truman, who presumably followed news media and who probably should have known even as merely Vice President and even though he could not use the product himself while still Vice President. (This is from various publications of history.) Without judging hypotheses by 20/20 hindsight, a claim in the 1940s that we could build a single bomb powerful enough to nearly flatten a city would seem to have been a fringe theory that not even the Japanese military commanders would have had reason to believe before one was detonated.

When the Soviet Union waged a war in Afghanistan and then Soviet veterans came home, some had post-traumatic stress disorder (PTSD) compounded by a popular belief that those vets had only served on some cushy military base within the Soviet Union and had nothing to complain about (I forgot the citable source). The Soviets apparently did not do much domestic publicity across the U.S.S.R. about the war; I know they did very little justifying overseas in English before it began (I read Moscow News, an official outlet, for the few weeks preceding and following the war's start as it took a few weeks to arrive in the U.S.). Thus ensued the popular initial disbelief of vets with war trauma. In other words, a claim that their own government was at war in a neighboring nation seemed to qualify as a fringe theory back then. Yet, according to The Atlantic, nearly 100,000 Soviet soldiers were involved against 90,000 Afghan Mujahideen, some cities and highways were put under Soviet control and some villages were flattened, about a million civilians were killed, and the war lasted over nine years. At least U.S. news media reported on it at length. This illustrates the possibility of a gray zone where information might leak to people who are not supposed to know but if they know the adverse risk or effect is low even from the viewpoint of a secrecy proponent.

Professions seem to vary in the likelihood of keeping a secret and people isolated from the people they might leak to seem less likely to leak.

It would seem more factors have to be included in the formulae. I'm not an expert, but I wonder if some version of game theory could contribute. Some of it relies more on experts to judge the likelihood of various actions by various actors and 100,000 or 190,000 would be a lot of actors, but actors could probably be grouped by similarity of interests.