Literally thousands of posts, and it would be impractical to repost them one by one.

On the other hand, anyone could volunteer to help with the repost, and more people can split the task, but the programming requires some skill level, and splitting the task further complicates it. Also, it is not necessary to restore all articles; we could start by the most popular ones and finish when the volunteers run out of interest.

But before proposing solutions we should make clear which goals are we optimizing for.

As readers, we want to be able to read the beautiful articles written by Yvain, and forward them to our friends. In this regard, some articles are more important that the others.
Yvain desires a relative anonymity, probably (my guess) for firewalling his professional life from some of the topics mentioned in his articles.

Hiding the articles from people who don't have high LW karma (or other technical obstacle) does not let me send the link to someone else. I think it would be better to have those articles freely available; just remove any traces to Yvain. And that cannot be done automatically, because some articles may contain personal information. So the exposed articles need to be checked by humans, and modified to hide the personal details, if necessary.

I am not sure about exactly what level of anonymity Yvain wants. Perfect anonymity is not available if the articles remain available; all the discussions on LW, including this very thread, provide connection between the author and the articles. But my guess (which has zero value unless explicitly confirmed by Yvain) is that he simply wants to prevent a possibility of his colleague randomly connecting him to the articles, using only five minutes of googling, without actually trying hard.

I mean, if the colleague is a LW reader, then Yvain is already exposed to them. But suppose that the colleague does not read LW, only finds a link to some "sensitive" article posted on their friend's Twitter; and he considers the article very interesting; let's say interesting in a wrong way -- he is so offended by reading the article that he thinks: "This scum deserves to be fired, because he violated my precious taboo!" Let's suppose the colleague does not work for NSA, but can use google. So he puts "Yvain" in the google, and keeps following the links... How can we prevent him from getting to Yvain's identity?

One necessary step in anonymisation would be to invent a completely new nickname, which would not be used anywhere else. Not even in LW discussions. Otherwise the link will be created retroactively. -- Unfortunately, I think this is very likely to happen; I mean, Yvain's first name was already mentioned in the very thread that discusses his need for anonymity. So even if we report his old blogs under new nickname, let's say "Papa Smurf", it is only a question of time until someone writes on LW: "I enjoyed reading 'Papa Smurf and his opinion on some social taboos'; John Doe is such a great writer!" Do this repeatedly, and it's a question of time until you write "Papa Smurf and his opinion on some social taboos" and get "John Doe" mentioned in one of the first three hits.