[daveguy]

Author mentions difficulty of reproducing research as one of the issues in passing. There was a recent Nature special edition on the subject:

http://www.nature.com/news/reproducibility-1.17552

I believe this is the primary issue and the cause is from one of two causes:

1) secret sauce in research -- details are lacking because there is a push to commercialize things that come out of academia

2) insufficient experimental design -- small sample size, poor controls, etc.

I would like to see an open publication that as a part of publication the result must be reproduced in a separate independent lab or two. This would almost double the required funding (maybe less because you eliminate false starts). Maybe just a few institutions could handle many reproductions.

There is a bit of a self-healing aspect in that the non-reproducible and non-interesting/advancing studies just get dropped on the floor. However, it would lend a lot of credibility to a journal that required an independent research confirmation.