I've always thought something like this would be a better way to approach public discussions, and their implementation seems to be fairly decent so far. There doesn't seem to be a whole bunch of activity yet, and not really much free style discussion under the few points I browsed, do you happen to know a particularly busy topic one could look at?
If it required you to make arguments in the form of First Order Logic statements and then did automated validation of them, then I would use it. I'll grant this is probably a niche use case. With the power of modern automated reasoning, and maybe automated fallacy detection, it seems a shame not to use that. Computer assisted argumentation would be a great forum feature.
That makes no sense in practice. The reasoning you employ on a day to day basis is completely outside the realm of what you propose. Realistic everyday reasoning is too complex; it is not satisfactorily represented by first-order logic.
Also, even mathematical proofs are cumbersome to state in a purely formal way. Moreover, many of our arguments hinge more on things like accuracy of assumptions, informal inferences, disagreement on "axioms" and goals, etc, not on formal validity of arguments.
This doesn't happen. Most of the time people either:
1. post something where you can understand what they mean but they didn't spend the time (not that you should have to) to refine the point and people are pedantic
2. post something so absurd that has no basis in reality.
I agree that the end result is screaming at each other but nobody is on Twitter to engage in debates in the first place. Nobody can present a valid discussion that was hampered in any serious way by twitter's character limit.
Unnecessarily simple arguments -> People get upset because you're misrepresenting them -> essentially screaming at each other instead of discussing.
The fundamental nature of the platform is bad. It's basically arguing in clickbait form.