I’ve just seen this in Google on an experiment an acquaintance of mine is running (she doesn’t blog on SEO, hence it being here; she also OKayed me writing this up). A recently registered domain, without having any links pointing to it, is now indexed.
There’s private registration, although because she didn’t immediately “activate” private registration, with Namecheap, it’s possible to see the old registrant info prior to activation. The person who registered it has some other trusted/well-ranking sites, so it’s possible that by correlating the whois info, Google is expecting that this will be a quality site and decided to index it.
Myself being a Namecheap client (thanks Smaxor for the free SSL tip), I decided to speak to support. They don’t autosubmit domain names to Google. So submission is not the reason the domain is indexed.
What’s also interesting is that when this domain – a nonsense keyword domain – was registered, there were 0 results for the madeup keyword. Now there are a few besides my contact’s domain; she and I have searched for it a couple of times. Perhaps Google [supplemental?] indexes pages that it wouldn’t otherwise index when it realizes they have unique keywords on them that are being searched for. Of course, that could easily lend itself to abuse, which is something Google’s engineers are weary of when deciding if a criteria is worth using.
A final possibility I considered is that there’s a trojan on my friend’s computer or mine. Keystroke logging might then have got spammers to find these keywords and autogenerate the junk constituting the rest of the SERPs [which, again, weren't there when he decided on that nonsense keyword]. I think this is the least likely of the possibilities though.
Update: My friend Andrew Shotland has a post featuring over a dozen techniques to get indexed without links or submission!Algorithms, Case Studies, Domains, Google