This is good to get some ideas and thoughts brought up...the drawings that Dixiedevil posted did say they were initial concepts. Like dielectric said, the rubber may not be the swiftest solution - perhaps the clamps collars are machined for particular frame size. After all that is just a pass of the ball nose cutter. However, I do think clamps would work. Do a quick look online, and even on this forum, for people who have used frame clamps from dennis kirk, or even mtb bicycle stem clamps. Here is a quick example:
http://dotheton.com/index.php?topic=6149.msg67826#msg67826
If one went as far as knurling the inside, I would be amazed if it caused frame failure due to stress - if this was the case people handlebars, or even their triple tees would be snapping left right and center. If a clamp can hold the force and weight of a motorcycle on a polished fork leg, then it should be able to hold your leg resting on the peg. Also, the forces to pull/strip the bolt from the 6061 is quite considerable. I've used Aluminum rearsets on racebikes with smaller bolts and no issues. Perhaps, and correct me if I am wrong here Dielectric, you're concern is stripping the threads from torquing down the clamps a huge force (to presumably keep them from slipping). The clamps would not need that much force, but I can see how that would, and is, an concern well made.
Dielectric does bring another good idea with the welding option. While the aluminum can't be welded to the frame, there is nothing to say that some tapped steel bosses can't.
Oh..and the bonding thing.I have more faith in a mech. clamp when it comes to metals , but that's a personal bias and subjective at best. I've found that strong bonds are either really brittle and can't handle vibrations well, or if they are designed for vibrations then they are pretty elastomeric (which would create similar issues to the rubbver shim/sleeve). If bonding can perform as well or better, then they speak for themselves. I'd be interested to see the specs on that stuff and it could be a great and clean way to do it. We have some really interesting bonding solutions we've used in our engineering department for various problems, so I am all ears to hear about new stuff.
Hanly has a neat idea with the pinning. A pinned arm with a tension plate to hold the angle could reduce the amount of material needed. I'd probably want to put a press fit sleeve into the hole to help with the forces and the frame. There are press fit inserts for sportbike swingarms, so I am not too worried about a press fit insert placed in one of the downtubes of a heavy steel frame.
Discussions and ideas amongst people help refine a product...so let's keep tossing some thoughts dixidevil's way