This is a bit unfair to the public. With any science, in my experience there are naturally different levels of disclosure. It is natural to share first with current collaborators, then with friendly recent collaborators, then with colleagues in related areas, then in informal seminar discussions, and so on.
One of the reasons for this is trustworthiness, as you say, but there is much more to it. Trustworthiness isn't the main parameter at all. A larger part of it is how much you trust the results. At each level of disclosure, by social convention there is a rough level of implied confidence that can also be spelled out more carefully. It is embarrassing to be wrong, and the higher your implied confidence was the more embarrassing it is! The rationale behind this social convention might once have been trust, but it is now free-standing.
Some incentives for layered disclosure are obvious; for example, for PR it might be useful to build a buzz within your research community. Other incentives are less obvious. For example, you might disclose early results to a colleague to hopefully get them interested and involved in the research. But if you disclose the results too broadly, too soon, then they will probably be less enthusiastic. They want the inside track on a cool problem, so that there is some time and freedom to explore it, and it isn't a race. (Races for results are stressful and also unfulfilling and wasteful; there's usually no point having multiple teams competing to do the same stuff, just trying to beat the others out for priority.)
One of the reasons for this is trustworthiness, as you say, but there is much more to it. Trustworthiness isn't the main parameter at all. A larger part of it is how much you trust the results. At each level of disclosure, by social convention there is a rough level of implied confidence that can also be spelled out more carefully. It is embarrassing to be wrong, and the higher your implied confidence was the more embarrassing it is! The rationale behind this social convention might once have been trust, but it is now free-standing.
Some incentives for layered disclosure are obvious; for example, for PR it might be useful to build a buzz within your research community. Other incentives are less obvious. For example, you might disclose early results to a colleague to hopefully get them interested and involved in the research. But if you disclose the results too broadly, too soon, then they will probably be less enthusiastic. They want the inside track on a cool problem, so that there is some time and freedom to explore it, and it isn't a race. (Races for results are stressful and also unfulfilling and wasteful; there's usually no point having multiple teams competing to do the same stuff, just trying to beat the others out for priority.)