A few thoughts on this after some research, being a self-styled "minor machinist" and interested in metals:
(1) About "shallow threads." Machinists generally aim to cut threads with 50% engagement on steel; threads on aluminum and brass often formed around 75% engagement. If you look at the threads (brass or aluminum) on your stove, you'll notice the roots are deeper, whereas the roots on your canister are shallower. But I'm not convinced this is necessarily bad, or out-of-spec. Plus, I'm almost certain the threads on the canister are rolled (and I believe rolled threads often have a shallower minor diameter).
I have a hunch the primary issue with thread failure on the stove is simply the steel canister vs. brass or aluminum, mixed with abrasives (dirt) in the threads…obviously, over time, brass, and raw aluminum, would wear faster than the steel. If thread failure were an ongoing problem, you'd expect to see a steel insert used on stove threads. That would be a low-cost, low-weight option, so their absence makes me wonder if thread failure on stoves isn't rare.
This also makes me wonder if the idea of lubing the threads with silicone grease is good or bad. That would almost certainly cause contaminants (dirt) to adhere to the threads. I wonder if a good blow-off, followed by a simple wipe of the threads with the bandana wouldn't be best.
(2) What I haven't figured out yet is whether there is a brass vs. aluminum advantage. Brass is harder than aluminum in its raw state, and has more lubricity, but the aluminum is typically hard anodized, which can approach steel ranges in hardness. Researching this, I've found multiple anodizers who state a good aluminum hard anodize can reach an equivalent on the Rockwell C scale of 50-70, which is very hard. That's knife blade hard. Of course, it may only be 4 mils thick or less (which is why you'd never see hard anodized aluminum used as a knife–sharpening would take off the hard layer). But that hardness should make the stove threads quite wear resistant.
Brass is measured on the lower Rockwell B scale, usually 55-90B, but even 90B doesn't reach the lower numbers of the C scale. 94 on B scale is equivalent to 15 on C scale. So brass is *considerably* softer than hard anodized aluminum. And yet, brass seems to work just fine most of the time. Calling into question again whether hardness of the threads is the issue.
If anodization were to "flake off," that would represent a failure of the hard anodize. That shouldn't happen. A hard anodize should wear away slowly and gradually, not come off in chunks. Because it is not a treatment that is resting on the surface, like paint. A good hard anodize achieves about 50% penetration into the surface, and the other 50% builds up on top of the surface. So if you see chunks of anodization leaving your stove, I'd say that's a good reason to warrantee it.
Torquing down hard on the threads is a different issue–hardness of the outer skin of aluminum would not protect it from deformation. Now we're talking powerful wedging and deformation action on the threads, and I can imagine that causing a thread fail of either aluminum or brass in short order.
So thinking this through from the perspective of metallurgy, the two killers of stove threads that seem most obvious would be:
(1) Dirt acting as an abrasive — so blow and wipe the threads before attachment–perhaps not lubricate with anything. And use the plastic cap on the canister.
(2) Over-torquing — because neither aluminum nor brass have great shear strength. In some instructions I see that the stove body should be finger tightened by the stove's base–the stove is tightened onto the canister, not the other way around. I can imagine if you held the stove stationary in one hand with a firm grip around the head or body, and torqued on the canister with the other, you could greatly increase torque because the canister has a much larger diameter.
