His post resonates very well with my thinking: an identity, even a digital one, is a lot more than "a set of claims made by one digital entity about itself or another digital identity" (from Wikipedia, which probably got it from the Identity Gang Wiki).

He proposes this mental experiment:

One way of sussing this out is to ask: do identity twins have different identities? We would say yes, even when we can't tell them apart.

The "even if we can't tell them apart" is the key phrase here. There is (digital) identity even without claims, and there certainly are different (digital) identities with the same claims!

He says:

Even a tech-savvy person like me has a hard time envisioning, never mind comparing, the interaction scenarios proposed by various identity schemes

I completely agree, and would add that nobody, not even the "insiders" really understand what consequences all the different proposed architectures have in terms of, say,

• who gets empowered and whose power diminishes
• how the attack vectors on those different architectures differ
• what governance structures are needed once any of this becomes a mass-market technology
• and so forth.

Which is why I think the Internet Identity Workshop put together by Kaliya Hamlin, Drummond Reed, Doc Searls and Phil Windley is such a good idea. If we all do our part, some real insights will come out of this workshop.