Art, maybe a demonstration of your thoughts by a practical example would be helpful. This way, we'd gather a better understanding of what you're saying while also producing an article that can potentially be used. Also, I am not disagreeing with you at all; I'm simply focusing myself on one aspect here (of which contributors have already submitted articles for) which is convincing practitioners that open source emrs are ok.
Sam, the sourceforge review structure has changed some: http://sourceforge.net/projects/openemr/reviews/
Note they are all 'YES' now because reviews from anonymous users are no longer accepted on sourceforge. So, now we are at 100%
I think you do bring up a good point, but would change the wording a bit since it's really not the same as the proprietary acceptance rate you are comparing it too. For myself, this is only a minor point. Also agree "Buyer's remorse" is better than recividism. With this term, I think could actually keep them separate ('Buyer's Remorse' section above 'Acceptance Rate' section) since it's much more clear. I'd agree with yanking ROI for now.
Also Sam, I completely agree with your Windows xampp package assessment; this is something that is really needed for OpenEMR 4.1 .