The data also appears to be bogus (again)
It took me a while but I also got hold of this list of passwords. Do note that the source of the passwords is still undetermined. One source says it came from Writerspace, another source claims that it was undisclosed and now ElReg speaks of an Australian ISP? Vague, vague...
While on their twitter account it seems (I don't do nor want social website nonsense) they claimed that it got from "their collection".
But lets assume that the source was indeed a website or ISP. What website or ISP would keep invalid addresses in their database? Look at the list yourself (its not that hard to find; once again distributed in plain text even). "hotmail.com.jj" anyone ? Or what about 1 e-mail address which appears multiple times with multiple passwords ....in the same section ?
It gets better.. I simply tried to verify a few addresses myself. Not by sending e-mails, that's spamming and intrusive. By using either the "verify" SMTP command or simply using RCPT TO: and check the results (after which I cleanly disconnect). Most tested valid but a lot also appear to be invalid.
How does any decent ISP or modern website (e-mail verification is so common these days) end up with non-existing (hotmail.com.jj) or invalid e-mail addresses ?
wrt the sections: there are three; one which lists "<password> | <email> | |". One which lists "<email> | <password>" and finally one which does "<number> | <password> | <email>".
Australian ISP? Its possible I suppose, but when looking at the list you'll notice that the majority are accounts from Brazil.