In an earlier post, I commented off-handedly that science is done by people, not institutions. Elsewhere in the blogosphere someone took exception to that. So it seems a point to expand a little.
The thing is, science is done by people, and scientists as people are rather heavy on individualism. Managers of scientists usually wind up making comments about 'herding cats'. Consequently, when you're talking about an institution that scientists are involved with, it's a near certainty that some of the people there disagree to some extent or other with what their colleagues published. Plus, as I mentioned in talking about peer review, merely getting published does not mean that you're right.
The scientists simply do not are not speak for their institutions unless you're told emphatically otherwise in some specific case. I do not speak for my employer, someone at NASA does not speak for NASA in writing a scientific paper, someone at NOAA does not speak for NOAA, someone at a University does not speak for the University, and so on.
There's a different aspect of that individualism. Even where the scientists are naturally inclined to be more conformist (which has struck me as rare), the rewards are set up for nonconformity. The surest way to prominence in science is to support an argument that what people previously believed to be true, isn't.
Between the two, by the time you can get scientists to agree overwhelmingly on something (the earth is round), it has ceased to be a scientific question. The shape of the earth question is now about things like whether the equatorial radius is 6378160 or 6378240 meters. (Maybe a narrower difference now; it's been a while since I looked.). There's agreement about round, but not about exactly how large, nor exactly how much squishing there is towards the poles.