I had to take a break from my researching, and I've probably forgotten most of what I'd come across, but I'm pretty sure it was evident in the thread that having the bios reroute the traffic ultimately seemed to make more sense than modifying every xbe that requires access.
With Live blocking being a standard on just about any bios (as well as nkpatcher), it seems logical to think that rather than just block the traffic, attempt to re-route it. I don't know if that'd be a more difficult way of going about it or not since patching xbes has become a pretty standard practice, it might just be easier to go with that.
I'm sure the guys that handle Kai/XBC would have a better understanding of the networking issues we're up against, but it was kind of tough getting anyone on board for the project because no one really wanted to attempt an XBL hack. My whole theory was that we weren't going to be hacking the actual Live service, but I couldn't manage to get the support I was looking for, and I don't have the programming/networking knowledge required to do it on my own. Good luck with this. Maybe we'll see some results in the near future.