Thanks, fellas!
So... after scratching my head for a while, I came up with a simple flat-ring design to hold the headlight--but it's going to be the default if my slightly more complex but more elegant-looking, experimental version doesn't work tomorrow.
I remembered the wooden die I made to form the indentation for the ignition switch in the R/H side panel. So, wondered if I made an oak female die to receive something sufficiently hard with radiused edges that's the right diameter, perhaps the thinner aluminum might simultaneously form and shrink into a compound-curve bezel, rather than puckering like a bottle cap around the outer edge (I have just enough thinner aluminum left for one try).
So, per the photo, I adjusted the template I had made for a simple flat bezel--widening the ring a touch to allow for a nice compound curve--and hunted around for a male die... voila. It looks like a 5 lb. plate from my small (not olympic) weight set is almost perfect.
The only deviation from the sequence I scribbled on the sketch is that I'll probably cut a 1" hole in the center of the aluminum sheet to receive a socket or something that will keep the aluminum and both dies aligned--so, a corresponding hole will go into the oak to anchor the centering rod.
I think I can route a perfectly round hole that corresponds to the diameter of the plate before its radius begins, then use a half-round bit that's lying in the drawer to mirror the male die's profile (roughly, anyway) and achieve the right O.D. for the bezel. Then, mate that piece to another piece of oak to produce a 3/4" deep female die with radiused sides, a sharp divergence from the bottom to the radiused sides, and a flat bottom to receive the centering peg.
Finally, I think bar clamps will be wise to reinforce the upper layer of the female die across the grain, so it doesn't split from what will be some significant pressure from the hydraulic arbor press--or I could screw the top to the bottom... or both! Because the clearance between the male and female dies will be just a few thousandths more than the thickness of the aluminum--and the metal will be forced to shrink as it forms--we're talking a lot of outward force being generated, I think.
Therefore, either I get a really neat, compound-curve aluminum headlight bezel with integral tabs at 3 and 9 o'clock for vertical adjustment and two others at 12 and 6 o'clock for the mount's second U-shape component to grab the headlight... or the aluminum puckers... or there's a small explosion in the shop and cast iron shards and/or oak splinters fly everywhere.
Definitely be wearing my face shield and some heavy welding gloves...