I agree strongly, conferences in particular are very important in this regard. Perhaps in some fields people are cautious about sharing full results at confere

I agree strongly, conferences in particular are very important in this regard. Perhaps in some fields people are cautious about sharing full results at conferences, but in most it is very much encouraged and beneficial to get your ideas out there to the community before a paper can come out. The "12 month time-lag" cited in the article usually includes at least one conference presentation, in which the main results can be presented, often receiving useful feedback that can result in a stronger finished paper as well. Could scientists make better use of modern communication media? Sure. I particularly wish more fields would adopt the arXiv model. But the peer-reviewed journal is not going to be displaced anytime soon, the best we can hope for is to make the process more transparent and open, more reflective of the interests of Science and scientists, and less of the for-profit journal industry. I highly doubt that the problem will be solved by "science startups." There are large non-profit interests as well, I expect the progress to be made by individual scientists, universities, and professional organizations who have an interest in destroying the status quo.