I can see some benefits of this but I think the idea of a script automatically going through and making changes is quite horrible. In one example they removed a

I can see some benefits of this but I think the idea of a script automatically going through and making changes is quite horrible. In one example they removed a new paragraph break. That's an unacceptable modification to me. I think that the best way to do something like this (if you want to do it at all) would be if the original author chose 30 small acceptable changes themselves. The presence of these changes (of lack thereof) would then allow a 30 bit ID number to be embedded in the document allowing each copy to be identified separately. They could also have a built in checksum to allow the original ID to be recovered even if some of the changes were found. By increasing this to 100 possible changes (50 active changes in each copy on average) then you could have a very large amount of error correction data and the original ID could still be recovered when a large number were found and changed (this would prevent pirates buying 2/3 copies to find the changes as 2 copies would only highlight half the changes on average if there is an equal chance of the change and original showing). Ultimately I do not think this will work however. A pirate can use a pre-paid credit card to buy the e-book. In this way they cannot be tracked. So I honestly don't think this is worth the effort. Just my thoughts on the matter.