The subpoena wasn't just for the public contents of tweets; it was also for any and all user information for the account, including email address and an anything else twitter might have collected, including IP addresses, physical location, browser type, etc.
In addition, the court ruled (yet again) that the defendant had zero standing to challenge (or even know about) the subpoena since it was directed at Twitter not him. Which is kind of terrifying given the push to cloud storage for everything.
Most of this fruitless discussion seems to be between people who have just read the TPM article (or scanned to the pull-quote from the judge), and those who are more familiar with the case. Here's an ArsTechnica article that communicates the issues quite clearly: http://arstechnica.com/tech-policy/2012/07/ny-judge-compels-...
The subpoena is for all public and private information Twitter has for a period of three months and a half. The charge is for disorderly conduct at one point during this period. The EFF and Twitter are pointing out that the prosecutor is going on a fishing expedition. The “shouting from a window” analogy the judge made is irrelevant.
In addition, the court ruled (yet again) that the defendant had zero standing to challenge (or even know about) the subpoena since it was directed at Twitter not him. Which is kind of terrifying given the push to cloud storage for everything.