Orin Kerr has written an article about "Digital Evidence and the New Criminal Procedure" and has asked for comments. In general I think the article is excellent, but I think Professor Kerr is missing an important facet of the issue. I sent him the following email (slightly edited).
I'm not a lawyer, so I hope that I didn't miss anything in your article that already addresses this issue: namely that in many cases it will be impossible to recognize digital evidence even when you see it, particularly when encryption is in play. (I did a search in your article for "crypt", just to make sure, and didn't find a mention.)
I wrote about this issue in the context of ownership and intellectual property here:
Essentially, any digital file is just a sequence of ones and zeros, and any evidence a file may contain depends heavily on how those bits are interpreted. The same file could theoretically be a grocery list when loaded into a text editor and a diagram of a bomb when loaded into an image viewer. Further, when a file is properly encypted it is essentially indistinguishable from random noise. Even if a judge were to issue a warrant for the decryption key, there's no way for anyone to prove that the file actually is encrypted and not just a bunch of random numbers.
In general though, I like your paper a lot. As everyone in society gets more tech-savvy I expect law enforcement will be able to work more precisely as well.