Email developed from a file transfer based protocol to the widely used SMTP (which was created in 1982). As a protocol it leaves much to be desired: senders can be forged, it's a playground for spammers, the ability to send anything other than plain text (ASCII) messages had to be added with duct tape at a later date, its error messages are cryptic and it has few facilities to deal with the complexities of back and forth exchanges of messages on the same subject. But email has one redeeming feature: its forward and store nature. The very essence of email is the ability to send a message, get it delivered to the recipient and allow that recipient to read the message when they choose to.
Instant messaging (starting with services like talk on Unix and other systems) are similarly poor. They allow two users (and sometimes more) to exchange text based message in real time. In general they do a poor job of doing anything else. Sending files or pictures often ends in disappointment as firewalls between recipients block the transaction. But for all its poor functionality IM is wildly popular, precisely because it allows simple communication in real time.
The same goes for SMS messaging. It's too short, character based and difficult to type (yes, Steve, even on an iPhone). Yet it took off. Although it is credited as having taken off because SMS charges were much lower than calling between mobile phones in Europe, I think there's another reason. SMS allows the recipient to decide what to do with the message. In that way, SMS is a polite technology. Unlike the telephone that interrupts the person answering it, SMS allows the essence of communication to get through quickly. (It is any wonder that MMS isn't similarly popular? If you send me a picture of your child doing something cute, we almost certainly want to have a conversation about it to satisfy human needs for affirmation.)
Skype is poor. For a long time I tried to use it for business purposes and it was too unreliable. Yet it's very successful at allowing people to talk cheaply for personal calls. The secret of its success, IMHO, is its ability to make the firewall problem that plagues IM go away. Even though calls vary wildly in quality, the user interface is difficult to use (and Skype insists on adding functionality that distracts from its core purpose), solving the firewall problem is its key value. Doing that made free voice and video calling possible.
Twitter is much derided, yet very, very popular. Why? It serves a need: asymmetric communication. Unlike Skype, IM, SMS or email, Twitter allows a person to graphcast: to send messages to people who have chosen to be attached to them. This functionality didn't exist in any previous form. This turns out to be very useful for marketing, and very useful for casual connections between people (without the people agreeing on the connection). In fact, Twitter shares an important characteristic with the web: it's permissionless.
Similarly, the web is rubbish. The web's greatest problem is that links rot, yet, in fact, that's the web's greatest triumph. Without trying to enforce connections between sites, the web allowed anyone to publish whatever they want... fast. Other competing hypertext systems were too complex and required too much central control to be successful. Twitter and the web work because there's no asking permission to follow or link.
Which brings me to Google Wave. What's Google Wave's single greatest piece of functionality that solves a real need? I can't see one. Google Wave looks like exactly what the designers said it was: "what email would look like if we designed it today". The trouble with designing systems after bathing in the mineral waters of Silicon Valley is you get functionality overload (as an aside, this is what makes Apple so successful: it's not what Apple puts in their products as much as what they leave out).
I wonder what the telephone would look like if it were designed in Mountain View today. "Well, Mr Bell, I'm sorry but your invention will never take off: the person on the other end is indistinct, I have to remember a number to get to them, I can't tell if they are even there to answer the call, I can't see them when I'm speaking, and how are we going to exchange pictures? Sorry, come back when you've invented telepresence".
So, next time you are inventing a communication technology (or, possibly, any technology), ask yourself: "Is this crappy enough to be successful?"