Thanks for bringing this here. It's a good idea because I would like to discuss it and have been going back and forth on whether I should post or not! Apologies in advance for rambling. I will try to actually get my main points across.
I actually think COIs can be problematic. We know, through research, that something as simple as a free pen or pad of paper can cause someone (I think the study I'm thinking of looked at medical doctors) to view a business more favorably. We also know that this can influence people without them realizing it. So I do worry about loss of objectivity and think we need checks and balances to counteract that.
And we have checks and balances. We have informal peer-review through colleagues (which is often the most brutal!). We have formal peer-review. We have editorial boards. We have the published results available for interpretation. We have replication (or lack of replication) of results. We have real-life results. Now I don't think the scientific method or peer-review process is flawless, but I do think it works well most of the time.
I think the study kathymuggle posted is somewhat important to address. http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2758529/#APP1 There's unfortunately not enough information in it to really do a good job and it is not specific to vaccine research, but it stood out to me that some of the things the authors considered a compromise of presenting results are not things I would consider a compromise (such as delaying publication - could be, but more often isn't). And the authors aggregated their compromises to make things look worse than they are. Having said that, I do think 15% of participants being asked to more favorably present data is potentially a huge deal. And the study she posted is not the only one highlighting the potential effects of COIs.
As for other things often pointed out as COIs, many of them I think are actually good things that require additional oversight, but are not in and of themselves bad things. Industry funding research? Yes, vaccine makers should fund the research pertaining to their products. Revolving doors? Yes, I want people who know the science, know the industry, and the know the people involved to be in regulatory positions. They know what's going on and can most efficiently and intelligently make the system work.
But I suppose the question still stands ... we do know that pharmaceutical companies and researchers and journal editors have behaved badly in the past. How often do all of those things come together and result in harm?