Obiously if you had nanotech you wouldn't go for a fucking chrome arm. It does deal with the problem of surgeries and the open wound thing, and how old implants may become out-of-date pretty fast. It replaces a surgical procedure with something more akin to upgrading your smartphone.

The interesting point about the brain sensor article wasn't the melting away - it was that there was no immunoresponse. It'd probably happen if you waited a bit longer, but we do have materials that don't inflame immediately. Which shows that there's still progress to be made.

I'm also surprised that you claim "any implant site is subject to opportunistic infection and everything else has a much faster evolutionary cycle.". What about hip implants? That's a fairly huge metal thing in your thigh. What about cochlear implants? (Or breast implants) They stay in your body, sometimes for your whole life.

Saying that we do not have useful implants is just plain wrong, we do, and they are routinely used. And they improve people's lives. They might not make people better at something they did before (except maybe breast implants?), but I doubt it's gonna stay that way for long. In fact, runners with prosthetic legs can already be faster than normal runners (which is just a narrow employ of legs, though, so in my opinion that doesn't count).

We can avoid the immune response on two ways: Suppressing the immune system, which is probably a bad idea.

The other idea is, of course, to avoid drawing the immune system's response altogether. And we can already do that, by coating it with biocompatible materials like collagen or PEG. The problem is to keep the body from reacting with the implant itself.