RF-ID uptake slower than expected?

I've been a quasi-believer about RD-ID technology for a while now. I know more than several people who are investing their time and money into this space, but I am a fence-sitter. Briefly, why?

• From what I hear and read, it doesn 't quite work reliably yet. The readers can easily be confused by multiple tags, mis-counting or mis- identifying.

• While the advantages over bar-codes are very real, they don't strike me as real enough to effect the kind of rapid industry change that fans of the technology are betting on.

• The backbone infrastructure that will be needed to process all this super-detailed new data doesn't exist yet. It strikes me that it might be an effort of Y2K proportions to upgrade the overall Enterprise software stack (ERP and others) to accomodate it.

That said, while I am a fence-sitter, I do think that RF-ID technology is very important and represents a real sea-change. I just think that it will take much longer than people think.(Remember one of my standard nuggets, it's easy to predict the future, what's hard is predicting exactly when.)

I was interested to see this article in the New York Times today about Wal-Mart delaying of their oft-cited deadline for all their suppliers to adopt to RF-ID.