This reply needs MATH, but more like the statistics branch. It's a mathematical function, that never goes 'catastrophic'. Or like an amplifier circuit, that doesn't go unstable, like some can.
Ah! You have the exact same same situation physicists found themselves in, a bit over a century ago, when examining black-body radiation. Simply put, their models turned out to predict that thermal radiation (black body radiation) is infinite, something that Paul Ehrenfest called
ultraviolet catastrophe in 1911.
The solution then, and to your question, is
quantization: instead of treating the statistics as a continuous curve (like classical physics did to thermal radiation), you note that it is actually non-continuous, consisting of small packets, quanta.
The tentative theory says, you can keep adding one individual after another, that is easily duped as mentioned already. But never reach a tipping point, where continuation is impossible.
This is because you forget that social interaction includes the occasional expression and application of "common sense"
by individuals.
As an example, consider
avalanche breakdown (or more generally,
electron avalanche). Say, you have a high-voltage source, connected across some thin insulator, and you slowly increase the voltage over the insulation, also measuring the leakage current.
A classical/simplistic model would say that the leakage is zero until it suddenly isn't at some voltage. That is not what happens in reality, because you do start to have some tiny leakage currents at relatively low voltages, which then increases first slowly, then in an exponential manner close to the breakdown voltage.
Humans behave the same way. Before and after they join a venture, they discuss (often socially) about it with others. Initially, especially in a successful scam, this is what lures more people in. Humans think much more when they talk compared to when they listen, which also explains things like the effectiveness of
rubber duck debugging (explaining your thoughts to a rubber duck also organizes your thoughts, often helping you realize/solve the problem you are thinking about). This means that as individual humans start talking about it to others, they also start sporadically applying their filters. This affects their output, and this spreads among the participants in the venture. If it indeed is a scam, at some point this effect has the same kind of "knee" that electron avalanche has, starting to feed itself almost exponentially, until the social atmosphere toggles from pro-venture to anti-venture.
My start in this dynamic was, in thinking of a daily personal task list, along with medically active 'fatique'.
I understand. For a few years, I had to avoid creating any task lists, because even the perception of undone/waiting tasks made me freeze/block/lock up, and give me insomnia.
The solution is to prioritize.
You use one main list that has a very limited number of slots. For me, this is a small wall board I put post-it notes to, one note per slot. I recommend against using your fridge for this, because for me it upset my stomach and lead to digestion issues. I use a wall in my vestibule: easily seen and noted, but not always in my line-of-sight.
You also have a wish list with unlimited number of slots. For me, this is a box of post-it notes.
Before you can add a new task to your main list, you must either complete one (so it can be removed), or move the least important one to the wish list.
Whenever you manage to do a task, or your priorities change, you go through your wish list, and move tasks between the main list and wish list.
If you keep a diary of your medical/mental issues (which I do recommend, because it is invaluable help to your health specialists, and makes it much easier to e.g. adjust your treatments and medication), you can use that diary to track and adjust the number of slots in your main task list. When you are doing well, you add an empty slot to the main list. When you feel overwhelmed, you move the least important task to the wish list and discard the main list slot.
I started with a single main list slot myself. Now, a few years later, I no longer get overly stressed about the existence of the main list and its items, simply because my view of them has shifted from "these are things I must do" to priorization, to considering "how important is this" for each matter, and weighing them against each other.
Which, if you think about it, is the most rational way to deal with multiple things you need to do!