Silver in comment #4 says:
it’s not at all clear to me why a “single universal personal identifier for multiple communication services” could not be their existing website.
Why not an existing website? Well assuming you're using standardized formats so the info can be correctly parsed, why not? Except of course that using http + parsing hCard or other formats is incredibly slower than doing a .tel DNS lookup. Plus, you need the skills so you don't make a mistake in publishing your data. And finally, you need to have a website already. Believe it or not, there are as many companies without websites as companies with. If I'm a window cleaner, plumber, dentist or doctor, why do I need a website? I need a Presence Online, I need to Market Myself, I need to Be Found and Contacted, but I don't need to waste money and time building a website.
Furthermore, even if I do have a website, updating my contact info when it changes can be a chore. If I also have a .tel, I can dynamically pull the .tel info and display it on my website, and use the .tel as an easily updated dynamic data store for contact info.
A simpler way of looking at this is to state that .tel is a business tool.
you need a site, you need to pay for it, and you need a means of updating it
Those are not dependencies for the .tel beyond paying the yearly domain fee (which you also have to pay with your own domain). You do not pay site hosting and other hidden fees, you do not build a site, and you have all sorts of means of updating it, from a standard web interface to complete APIs or even uploading excel spreadsheets.
Also I think you're missing the core point that .tel is (once again) not a website. You're saying:
I’m pleased to hear that .TEL may incorporate microformatting in the future! It definitely makes sense if you want to facilitate machine-readable contact info.
.tel is absolutely the easiest machine-readable contact info system there is! All .tel contact info is stored as structured DNS records. Simply make a DNS query to a .tel domain, asking for the contact records, and you've got them right there. They need minimal parsing, and probably about 100 times less resources than grabbing the data from an hCard. Navigation at the speed of DNS is something to be appreciated. :)
To go back to Yellow Pages directories: they're not competitors to .tel. They'll gladly grab your .tel info to populate their directories, it'll save them quite a few headaches and will make their offering more valuable. It's only that currently SIP and other communication channels are short-changed because of the difficulty of modifying the underlying infrastructures of the major YP providers.
Finally, let me answer your last question:
The privacy feature is the least clear selling point about .TEL. I’m by no means ignorant about privacy issues, but it’s unclear to me how .TEL helps those, and why that those of us who want to make our contact info more readily findable should simultaneously want to keep it private?
To start with, there's a comprehensive .tel Privacy document. You certainly want your contact information readily findable, but are you sure you want all your info findable by everyone? Probably not. You may want to keep that cell phone number of yours private and only give it to a couple of close friends. In order to do that, Telnic has developed a privacy model comprised of encryption specifications and an optional centralized friending system implementation. I've described the implementation in another blog post of mine, Privacy in .tel.
Here's a final thought for you:
Imagine you want to dial a number, but instead dial a name. This is .tel. The name resolves to a number (or many) and the call is made. It's just like typing and name in the browser address bar and it resolving to an IP address. We're in the 21st century. 15 years ago we stopped remembering IP addresses. Why do we still need to remember phone numbers?
PS: The .tel-to-vCard mapping info is here.