One of my managers had a rule of 2 (different from my rule of 2).

    You haven't proven you know how to do something until you've done it twice.

That was expanded by later managers to a more general sense of reproducibility, where 'we' don't know how to do it until two other people have done it.

There are certain aspects of my life that are informed by a general dread that some day I will leave my house and never be seen again. For instance I try to tell tell people I love them when we part, even if I'm mad at them, in case those are our last words for a long while, or forever.

If I really thought I was doing something at work that would forever change a corner of the world, I'd be super paranoid that the secret could die with me. That's the perfect recipe for me overworking. If I can't trust anyone with the information (eg, intellectual theft), then I positively vibrate. If I told people and they don't care, I have to finish it enough that I can demonstrate why they should care.

Yup. Like the saying I heard from a friend about how to treat engineering tests:

"The first test result is an error. The second is a coincidence. If the third lines up, then you might be getting a hint of a result."

As in: if you can't reproduce it reliably, you have nothing.

in science, if you make a prediction from a hypothesis, conduct a test, and what you predicted is borne out, that's not nothing; for example, if you look at historical records and hypothesize a reappearance of, I dunno, let's call it Halley's comet, and decades later Halley's comet reappears, why... they might name the comet after you!

Charts are for asking questions, not answering them. My life goes so much better when I remember that (or importantly, when I can find the right set of words to convince all the other decision makers that’s true).

An experiment gets you a little money and attention to try it a few more times. A few successful runs in different lab conditions should let you get a little money to start systematizing your experiment, and that should get you money and attention from industry, manufacturing.

That has always been an 8 year process at the best of times (most things take ten). Seems like a lot of people don’t have that sort of patience anymore, and they try to make things go from lab to shopping cart in three years or they endlessly loop on generating news articles (aka bullshit).

Precisely - these aren't "Warning signs" that are only clear in hindsight. They are proof of fraud beyond a reasonable doubt. If a person or organization has these traits and you continue to trust them, it is not an honest mistake.