The Dunning-Kruger effect isn't actually the same thing as overconfidence effect or the superiority complex I'm manifesting here, although there's definitely a relationship. What Dunning and Kruger hypothesized was that sometimes people are bad at an activity, like being funny, because there are important things about that activity that they don't know. This impairs both their performance at that activity and their ability to assess their performance, as well as, for example, other people's performance. In their experiments, they did find the predicted effect, with the result that, while people who were really terrible at an activity had a lower self-assessment than people who were pretty good at it, the really terrible people overrated themselves by a lot. They also found that the more competent people underrated themselves by a little, which was not a prediction of their hypothesis, and is the opposite of what the overconfidence-effect theory predicts.
At least, that's how I remember it. Would I even know if I were misremembering it?
I think "How do [I] signal a commitment to the falsifiability of [my] assertions and yet remain convincing?" is mostly the wrong question. Convincing someone, like a salesman, is a different activity from collaboratively exploring ideas with them, and they are not only mutually exclusive with one another, but also with the kind of self-righteous dominance discourse samhw is engaging in upthread, where the object is to persuade other onlookers to side with you against your contemptible interlocutor (in this case, me). Projecting cocaine-like irrepressible confidence is often a very effective way to convince people of things, because they assume your confidence must be well-founded, but that effect is poisonous to collaborative exploration of ideas, which involves looking for their flaws as well as their merits. It can also damage their confidence in you over time, although fraudsters like Lacan and that delusional coke freak Freud often get away with it for more than a human lifespan, by virtue of carefully crafting their theories to be nonfalsifiable.
In cases where your beliefs are based on objective evidence, you can share that evidence, as Darwin did with his studies of finches and reflections on pigeon breeding and whatnot. This turns out to be very effective at collectively progressing toward the truth even when none of the individual people involved is very smart (and none of us are) or very detached from their beliefs. Semmelweis was eventually convincing about handwashing, for example, but not about the cadaveric-particle thing.
Unfortunately, without computers, this doesn't work for procedural knowledge; every generation has to learn it from scratch by practicing, so, for example, our jokes aren't any better now than they were in Sumeria 5000 years ago. And there are a lot of cases where our knowledge has a basis that isn't objective, so we can't put it into words and numbers, with the result that people who trust us believe it, while people who distrust us don't.
Anyway, so people often misunderstand Dunning and Kruger to have found that the least competent people rated themselves as being the most competent. While that does happen sometimes, that wasn't what they found. However, sufficient levels of overconfidence in your knowledge will totally close you off to new information, as happened to samhw in https://news.ycombinator.com/item?id=28592769, so you totally stop learning, eventually making you the least competent. As I said, though, that usually also happens to humans when they go through puberty! In my case, my overconfidence stunted my learning about all kinds of things for many years and probably still does in areas I haven't noticed yet. And probably never will, since I'm apparently totally failing at actually thinking.
The "simulation of a smarter self" thing can actually work, because often what keeps people trapped is not a lack of intelligence but various kinds of mental formations they use to avoid discomfort. Doesn't have to be a self, or smarter; "what would Jesus do?" is a Christian reminder that a mental simulation of a more morally commendable person can be a useful guide.