Reuse from previous work, be it paper or blog, can greatly help reduce redundancy and make the relations between different efforts/results explicit. This will require

1. a good content publishing and linking model that allows easy traversal between description of different efforts without loss of context and attribution; to an extent, the web and blogs use such a model.
2. a community (not a select few reviewers) that is proactive in checking for validity of claims.
3. a change in incentives/yardsticks — interesting modification/extension to an idea should be considered as a contribution and as a small publication.

My 2 cents :)