I've gotten just a few too many of these kinds of these DOI fail messages lately and it's starting to get old. Call me crazy, but when a DOI resolution still doesn't work after one month and there's not so much as a confirmation that the problem is being looked at (or that maybe I'm doing something wrong), that's no way to serve customers or build confidence.
As part of an ongoing project, I've found numerous examples of failed DOI resolutions. The problem is that the source of these failures is the links contained in the syndication feeds created by the publishers themselves.
In effect, these DOI resolution problems make it impossible to view the article described in a journal's own feed.
To my knowledge, the DOI resolution service hosted at http://dx.doi.org is controlled exclusively by the CrossRef organization. Perhaps this lack of customer focus is the price we must pay for tolerating monopolies of any sort as gatekeepers of scientific information.
Perhaps it's time to think of a different way.