This is an good point. I think there is a difficult balance to strike between open communication and adding confusing noise to the scientific literature. Partly for historical reasons, there is an expectation that published science is correct to the best knowledge of the authors. Since writing, publishing, and reading papers takes a lot of time and effort, there are advantages to this precedent. I work in physics, and I can tell you that if we published all of our half-baked and often wrong ideas, we would waste a lot of people's time, at worst sending people down blind allies that we would soon rule our ourselves. I suppose tying reputation damage to publishing incorrect or misleading results is then part of the incentive structure that keeps publication quality high. At the extreme, there's was one recent LK-99 paper that had an obvious glitch in their data, and instead of taking a bit more time to debug it, they just posted the paper and speculated about what was going on. If that's how much you're rushing, how do I know I can trust your data?

But there are costs to this. There are big gaps between what people discuss with colleagues and what gets published, and the is no forum to publish partial or negative results, except maybe conferences. Ideally published papers stay at a very high bar, but there are other forums to publicly share work in progress. In a way Twitter is becoming this.