You want to get back in touch with Joe that you knew in 5th grade? How are you gonna do that over SMTP/IMAP? Easy on Facebook.
Or think about it this way: your aunt wants to put out photos of her family's recent vacation for everybody she knows to see. Is she (if she can possibly help it) going to go into her email app (or, worse, webmail) and make sure she's got all the right names with their associated somewhat-cryptic (well certainly more cryptic than a Normal Human Name) email addresses selected from her address book? Her address book (if it's even populated at all) probably doesn't have 10% of the email addresses that she wants to send to. And 1/2 of those are out of date or mistyped.
There's still plenty of room for separate services that do things with more options and with a broader reach. But for the basic cases (which is a LOT), having a single shared service that nearly everybody's on (and on which they don't have to remember a username->Real Name mapping), is a tremendous asset.
The rest of your points were part of my point. "Real Names" vs "usernames" is a temporary inefficiency in the early services. It's a typical programmer's solution causing normal person frustration. No-one's making that mistake again. No-one's going to roll out, say, an email service without a profile page for each account or a White Pages feature. No-one's leaving out simple photo/file sharing, tweeting or blogging as an integrated service.
And everyone who isn't FaceBook has a huge incentive to support the first federated protocol that enables all that. (So they stand some chance at capturing internet traffic again)
Say, Google's Wave.
Does anyone think it won't roll out with support for a profile page, some sort of White Pages feature, simple blogging, twitter-esque status updates, etc? It's already got email/chat integration and simple photo sharing this early. Why in the world would they be testing photo sharing if they weren't looking to do much more than 'Email 2.0'?
I used to think MySpace was for people who couldn't build their own website. With Facebook's success I'm changing that view. It's about the social connections not simply self publication.
I'm certain I'm not the average facebook user, but that's the value add for me.
The big 'hook', the reason it gets so much traffic, is that people use it as a proxy for the wider internet. It's many services funneled through one account. (As opposed to having a gmail account for email, an msn account for chat, a yahoo account for games, a flickr account for pics, a wordpress account for blogging, a twitter account, etc)
But what keeps it relevant tomorrow?