Wikipedia's policy is to not allow original ideas or opinions, even if it's from the subject of the article. This is fairly uncontroversial compared to some of the other things, like restricting edits to logged-in users.
Perhaps edits could be signed with a person's public key (verifying that it was them posting), thus making that edit, in and of itself, the primary source that the [following] edit can cite? It makes sense to rely on our IRL trust network to verify claims inserted by anonymous IP addresses, but when an actual, verifiable identity comes into the mix, that process should be short-circuited.
Wikipedia generally discourages edits by the subject of the article anyway. This policy is designed less for the likes of Donald Knuth and more for the likes of the borderline-notable people of the world who might be tempted to use wikipedia as their own CV.
Well this argument is as old as the hills (or more accurately about as old as the first reverted edit on wikipedia) but personally I'd much prefer a wikipedia which only contains notable information to one which contains all sorts of random information.
Now, I don't run wikipedia, but thankfully the people who do run wikipedia happen to agree with me on this point and decided long ago that only notable and verifiable information would be included. I'm really not sure why people continue to argue the other way.
Besides, who's gonna verify all the information I write about myself?