Tough question. Let’s see how this could be approached in Web 3.0 and Web 2.0 paradigms:
Web 3.0 - assuming everyone would have the AKASHA dapp installed with an Ethereum/IPFS node running in the background.
At the moment we use the ethereum blockchain to store the IPFS hashes pointing to the content, so removing the “link to content” without any trace would be impossible. Even if the hash pointing to the content would be altered in the smart contract log, since the content is stored on IPFS nodes scattered around the globe it would be very hard (if not impossible) to enforce the deletion of a certain file across the entire network.
Web 2.0 - this is where things get a bit trickier as now we have central points of failure in the form of IPFS gateways for serving a static version of the content to people that do not have their own nodes running. Until the IPFS JS implementation is ready gateways represent a compromise for offering the content found inside AKASHA to classic browsers users.
In this case, in theory, we could be forced to stop serving content found at X IPFS hash from our IPFS gateway. However, nothing stops users to setup their own alternative gateways - in this case it would be a matter of readers to select which gateway they want to receive their static content from. So again, deleting a certain thing from the network is not enforceable but, at best (or worst depending how you’re looking at it), we’re talking about “hiding it” from visitors on our website. I think it's important to note that this wouldn't affect AKASHA dapp users as they don't rely on gateways for the content distribution.
That being said this topic is very vast and complex as it has many ramifications that are simply too far reaching to be solved by a small team like ours. In other words, at this point, we think that the “right to be forgotten” is something very hard to achieve with technologies like blockchains and IPFS but, as our community will grow, I am sure that new and interesting solutions will emerge organically as we’ll have more eyes and brains tackling this sort of complex problems.
At the low, blockchain level it cannot exist, but it could be "emulated" at the UI level... same as for regular web: the result can be filtered from some search engines, but not from archives, copies, foreign/custom search engines, private or governmental databases, dark web databases, etc.
An alternative search engine would still be able to expose the data in searches, removing the results from google did not "forget" the data. The only "protection" is that it is not economical for most actors to have a search engine / UI on the scale of google, or to archive everything.
It actually does not take any technical expertise to go around google's "right to be forgotten" removal for instance (just use an alternative search engine, or just use a regional google search page), and certainly much less than forking a blockchain client, adapting and compiling it.