Ah, finally! A recent post from somebody outside of the OpenID devsphere about OpenID.
During my tenure at AOL, one of things I always wondered about was: How do external consumers and outside developers look at our OpenID? When creating the enterprise transactions systems, I had the luxury of using AOL internal IDs and OpenIDs, but when a user accessed our systems from the Great Beyond, the OpenID was frequently the only available ID source for the person/entity. There were even a few public blogs from corporate OpenID developers about how great OpenID was, and how it was overwhelmingly popular with folks outside of AOL. So, whenever the requirements architects came over and handed me their requests, during my architecture sessions it’d be brought up whether OpenID would be the sole identity mechanism to deploy, or whether for robustness, we should provide it as merely an available option. I’d opt to provide it as an option, with extensive identity checking and correlation– under many circumstances, my systems would be looking for one and only one valid user identity.
So to see dissension among potential consumers is always fascinating and an excellent learning opportunity. When creating APIs and frameworks, particularly things that integrate diverse systems with varying scaling capabilities, learning about how my customer base wants to use my stuff and how they actually do use my stuff definitely contributes to the continuous improvement of my deliverables, whether they’re low-level APIs, app-level components, frameworks, finished applications, and/or app suites. 
N.B.
None of the info divulged above is a company secret.
The “somebody” I mentioned above is a fella named Kyle. Phun stuff. 
