I believe it is necessary to find a balance. When scientists make discoveries that could be really important they should share them with scientists as soon as they have good evidence. However they should make it clear that these are findings and not final facts. This way other researchers can look at the data try to get the results and find any mistakes.

What happened with the BICEP2 announcement is an example of why this is necessary. When they announced their discovery early it was exciting. It also caused confusion later on. This is because other studies suggested that the signal they found was probably caused by dust in space not by waves from the beginning of the universe. While announcing discoveries to the public can get people interested in science it is important to be clear about how sure or unsure the scientistsre

In general I think scientists can announce discoveries before everyone else has a chance to review them as long as they are honest about what they know and what they do not know and they are willing to let other people check their work. The BICEP2. Other major results like it should be shared in a way that is transparent and open to verification, by other scientists.

the early universe was filled with matter and radiation that were almost evenly distributed, but slight differences in temperature and density naturally existed. As the universe expanded and cooled:

Regions with higher density attracted more matter through gravity.
Regions with lower density lost matter to the denser areas.
Over billions of years, these growing density contrasts led to the formation of galaxies, galaxy clusters, stars, planets, and other cosmic structures.

In this view, the universe did not need to start out highly irregular. Instead, small natural variations, combined with the long-term effects of gravity and cosmic expansion, were enough to create the non-uniform universe we observe today, while preserving its overall large-scale homogeneity.