Decentralizing science may lead to more reliable results

Their lack of openness stifles innovation. Rather than carefully investigate new claims, guild-oriented "Scientists" reject anything that's not very similar to past results. This ignores what we've learned from The Structure of Scientific Revolutions and betrays the thing science is uniquely good at - separating truth from non-truth.

There's too much social pressure to conform, and this is tied up with university research, publications, and tenure. Compare the Silicon Valley culture that assigns no penalty to failure to the university model that ensures anybody investigating something potentially revolutionary won't be on the tenure track. That's no way to run a scientific revolution.