An article at Washington Technology by Ethan Butterfield raises many thoughtful points, some of them new:

... states are waiting for the Homeland Security Department to issue guidance that they need before they can revise or overhaul their systems and processes, industry observers said. It could be several months, perhaps longer, before any guidelines are ready.

"I wouldn't expect proposed regulations until, at earliest, the end of the year," said Cheye Calvo, transportation committee director for the National Conference of State Legislatures in Denver. "Final regulations, we may not see until the spring or next summer."

In other words, almost a whole year (of the three years) will be wasted before computer systems can be defined and planned. (It would be worse than foolish to spend money developing the computer and work flow systems without knowing what additional requirements (and redesign) Homeland Security will require.)

NCSL has suggested that the direct cost of implementing Real ID will be between $500 million and $700 million nationally. But that would only cover the new licenses and systems. The largest cost is likely to be in hiring more staff, training personnel and opening new offices, many of which were closed over the last decade as states realized the efficiencies of renewing licenses over the Internet. The final cost of Real ID is likely to be much higher, Calvo [of NCSL] said.

"Certainly, it's going to rise into the billions. I don't believe there is any question about that," he said.

"The question is: Can it be done in three years? Can it be done in 10 years?" Calvo said.

This is the first time I've seen discussion of the timetable, and whether it may take much longer to make this system work. A ten-year plan would be a dreadful idea, becuase all the technology available now will be completely outmoded by then!

The article disucsses the likelihood that HS will require biometric ID:

Facial biometrics would seem a logical security feature ... but other biometrics are available. ... State officials don't want DHS to choose one security solution for all states. They prefer trying different technologies with various business partners, Calvo said.

"The concern is if you have one way, a uniform way, it becomes very static, and it ceases to keep up with the innovations and the wrong-doers," he said. "The counterfeiters are innovating, so you need to have different technologies constantly emerging to deal with the problems."