M. Guedon has a good summary of why the current peer review model is limiting the spread and updating of knowledge:

Excerpt:

1. The present system is too rigid, too unwieldy to permit such small-scale, yet potentially crucial interventions. To make the proper corrections, one would have to republish and perhaps even go through the publisher if it is in print. The communication process is therefore limited or blocked.

2. There is a second type of difficulty: the present system of scholarly publishing relies more on a credential system and a co-operative system rather than on the intrinsic quality of individual intelligence and the excellence of the submitted text. One does not enter scientific or scholarly territories without showing the right kinds of references – diplomas, titles, names of institutions, etc. As a result, the scientific and scholarly enterprises work as a two-tier system where the authorized write and read and the others do not write and often cannot read because of economic barriers, such as high subscription prices and lack of affiliation to the right library).

3. To address these obstacles, M. Guédon touches on the granularity issue. The article is not the only possible model to contribute to scholarly or scientific research. This is even truer of the monograph in the humanities and, in fact, the article has superseded the monograph in most disciplines. He suggests that knowledge should be regarded as a conversation. People should freely be able to contribute to it. In the scientific community for example, moving closer to a wikipedia model could be the way of the future as knowledge would be made available to everyone; it can be created together, modified on a global scale, improved upon, and so forth. However, the argument of quality comes to mind. He counters that the present criteria for quality inherently rest on a hierarchical vision of society. When excellence is sought, the greater the number of minds involved, the greater the quality of the work done: the case of free software and some recent analyses of Wikipedia confirm this general rule. The greater the numbers of people involved in an issue, the better the answers are crafted. Consequently, the lines that separate the experts from the rest of society should be erased. We will always have experts in various fields, but to limit contributions to knowledge as a whole to experts only is to deprive all of humanity of its enormous potential for distributed intelligence.