The article mentions that the biggest impact comes from limiting the highest-entropy information. It wouldn't be hard to rank the various attributes by entropy

The article mentions that the biggest impact comes from limiting the highest-entropy information. It wouldn't be hard to rank the various attributes by entropy and then you - individually - could calculate the probability that you'd be uniquely identifiable. Seems like such work would be valuable.