(Water)Markings ? the "French" alternative ?

In one of my precedent blog post (In French) http://readingandraytracing.blogspot.com/2011/12/drm-lexception-culturelle-francaise.html , I evocated about marking as an alternative to DRMs, but didn't say much more than that, so here are some details about what this term covers when I talk about it. Please note that some call these "Social DRMs".

For me, and in THIS precise context, "Marking" is a technical mean allowing, through modification of a digital ebook, to add non-intrusive data, relative to the selling circuit, up-to (and including) the final "sell" to the customer, this information beeing non-trivial to delete. The goal for the retailer and the publisher to deter "piracy" and illegal redistribution.

A few points of interest :

Visibility:

Marking can be made visible, or invisible. The solution used by Immateriel (the most used in France) is to at least show parts of the marks, which has at least one benefit : it clearly shows the customer that the "copy" that he received was given to him personally, and not to an indeterminate community. However, this choice can have detrimental effects on the reading experience, as is the case when the visible indications appear at the start/end of each chapter, especially if they create incorrect formatting.

I would personaly be in favor of a smaller visibility, restricted to the start/end of the book. This would not, of course, go against invisible markis inside of the text. In all cases, the visible part would have the "dissuasive" effect, while the invisible part allows "tracking" a leak's source if necessary.

Technical : 
First point : DRM are in fact already watermarks, as they already closely tie the content with personal data. 
And the "most available tools" that get rid of DRMs, clean the files from the encryption only, not necessarily the entirety of the personal linkings. So if you're happy with DRMs, why whine with Watermarks ?

As all DRMs, marking is fallible, which is why I stated the goal was not necessarily for it to be tamper resistant, but non trivial to suppress. One can note though that since they are less intrusive than DRMs, hackers will have less motivation to crack them, and they should stand better the test of time.

An other more technical point is that while DRMs are defectuous by design (as you need to give at the same time the product and the key to read it), the reseller can keep slightly more data secret, which is slightly better.

Unintentional divulgation :
Here is for me the sticking point : in the case where the work is distributed, either inadvertently, or badly ckecked,  the reader could have problems for things that are not of his fault. For example if he looses his ereader, he won't necessarily think about going to the police with a list of the works that were on it whereas this would be the only "good faith" proof if the works where distributed...

Once again, Watermarks are DRM-lite, and while not widely known, this would already be the case with standard DRMed books.

Not available at big distributors:
This intermediary level protection measure is not widely used (and not among the "big" resellers"). So if you prefer them to full size DRMs, please ask for them in replacement ! Also, since the big-resellers don't DO watermarks, they almost always "upgrade" them to full-scale DRMs... Not really the best for readers, is it ?

Edit 2012/07/06 : I think the "watermark" version of protection is gaining traction, for example J.K.Rowling uses it for her Harry Potter books at Pottermore. But until the major resellers (Amazon/Apple) don't support them...