You say that like it's not a big deal. It's 8 am in Honolulu, one of the smaller cl cities, and there are 200 cars posted today. There's no way to search by year, blue book, mileage, etc, all of which are (generally) possible with the existing data. I think craigslist could be much more useful.
The biggest thing holding them back is their uncharacteristic resistance to mashups. I strongly believe they're working against their own aims (public service) in shutting down services that scrape their data.
To summarize:
Search is key. CL are a non-player in our market, but the CL equivalent cannot compete with us in automotive.
Why? High quality reference data, the kind that means you can't add a 1999 Ford Pinto CL if no such thing exists, and if it did exist we know it came with an iPod dock, is reasonably expensive to create and maintain and generalist just don't have the focus.
High quality reference data leads to high quality search, where you search for 2007 Ford X and get only '07 Ford X' or you search for late model BMW's between $25k and $33k and get exactly those, tends to deliver high quality buyers. Search quality and lead quality are inextricably linked. Sellers tend to follow high quality buyers.
I don't know how CL fares vertical for vertical, but I'd be quite surprised if in most markets they were the #1 player in each vertical.
It's not an aggregator, its value lies in the fact that it has tremendous brand recognition far more than the sheer volume of data it has.
The craigslist folks could sell the company in the blink of an eye. The amount for which they could sell it is continually increasing.
It's great that its so expansive, and search is ok - but sometimes while scouring it, you come up with little gems that make you go "oh wow"...
If you could create something that would give you more "oh wow" moments, I think you'd have enough of a disruptive technology to draw away some of their users.