This might prove effective for idiotic small time terrorists (which the UK security services seem capable of catching, but which don't pose much of a threat), but there are many obvious ways to avoid this being an issue:
- Don't use technology.
- Be involved with a big enough plot (or bad enough co-conspirators) that 5 years in jail is no deterrent.
- Store all data online, so it's not provably yours.
- Disguise the fact that you are using encryption.
- Distribute the encrypted material and/or the keys.
- Use some kind of "slow browser" with its own private key which decrypts pages on demand, but so slowly that bulk analysis of large volumes are impossible (cos you'd dump wikipedia into your content repository, to keep them busy and ill-educated).
...and that's with about 30 seconds thought and no motivation.
Ironically, investment by the security-obsessed content industry in stealth watermarking, the consequent development of stealth P2P technology, combined with the huge data volumes now possible both on- and off-line, is probably moving the game on to where the "police" won't even know that the target content exists.
Do the spooks have the technology to spider Flickr, YouTube etc looking for stealth-encrypted data in pictures of people's pets?
Ultimately, if someone has the means to build/steal and transport a nuke to a city (one of the worst case scenarios for which such invasions of privacy are presumably justified), avoiding detection by electronic means will pose zero challenge, and prosecution after the event will be irrelevant.