Bill,
I also believe that some applications are better served with adhesives rather than rivets, and attaching a plastic canopy to metal framing is one application that begs for such use.
Riveting has obviously worked well, but I am not comfortable with stess concentrations at each rivet hole when you could better distribute that stress evenly over a much larger (bonded) area.
I looked at the specs for SikaFlex and Lexel to see how they would work for canopy installation to the frame, but I also looked at E-6800. Keep in mind that these are all urethane based adhesives (no silicone) but the way they are formulated does make them different from each other.
The short list is:
E-6800 has 8 times more tensile strength than SikaFlex (3500 psi vs. 450 psi) and twice the Shore-A hardness (80 vs. 40) which should mean (I say "should" because I'm not a chemist) that E-6800 is 8 times stronger than SikaFlex (an obvious advantage if I understand correctly).
The big downside to E-6800 is that it has only 5 minutes tack time vs. 50 minutes tack time for SikaFlex.
What that means to me is that you could apply the SikaFlex onto the frame and then install the canopy and wiggle it into position.
But with the E-6800, it would be better to be inside the cockpit with the canopy in position already, and THEN run a bead of E-6800 along the canopy to frame bond area. Then when that's cured, you can take the whole assemble off, turn it upside down, and run a bead on the other side of the frame.
Of course, you'll have to figure out how to get yourself out of the inside of the fuse after running the first (inside) bead... OR just sit there for 24 hours breathing the urethane fumes while it cures.
(Hey, I'm just the idea guy. I'll let someone else figure out the details of the operation).
Another thought is that some people have reported cracking of the fiberglass lay-ups done on the bottom edge of the stationary front wind screen of slider canopies (after some time in service).
So, instead of using epoxy resin and glass cloth, why not instead do that lay-up by saturating (forced squeegeeing) the longer tack-time SikaFlex into the fiberglass roving and use that as the basis for the lay-up.
Once that's hardened, you can come back with a coat of SikaFlex OR E-6800 as a top finish to fare it in and make it smooth. The 80A hardness of the E-6800 especially is enough to lend itself well to priming and painting (it's actually recommended that it BE painted for even better UV protection), but it's not so brittle that it will crack like epoxy fiberglass eventually will because the urethane adhesive will better hold up to vibration and frame twist over time.
Plus, my experience with both epoxy glass resins as well as urethane adhesives has proven to me that if I have to choose between the two, based on how well either will "stick" to a metal surface, the urethane wins hands down.
Just my thoughts on the subject, and I am leaning very heavily toward the idea of bonding, rather than riveting, the canopy sections on my RV7-A slider.
The only concern I have is in how to be able to get a perfect fit without having to force it before bonding.
Of course, whether you bond OR rivet, you don't want to have any built in stress anyway so that's something you have to deal with either way.