Simply putting another check in place, even with sensors and feedback, may do absolutely nothing to change the failure rate.
True! You cannot simply check the same things twice. The checks have to be substantially different.
Disregarding evaporation and seepage the chasm between the dams will fill and then 0.01% will start passing through the second dam. Not 0.01% of 0.01%, but 0.01% of the total... which means 24,000 gallons a day in our hypothetical example. The failure rate is identical.
Well, the dam model breaks down at some point. Yes, if we are only talking about water seepage, any dam will pass the water because each checks only by the same parameter. When I pick up a cartridge, I cannot measure dimensions to a mil, but I can assess general shape and proportion, get some idea of weight, and read a headstamp. The chamber is possibly a substantially different check, verifying maximum dimensional character in all three dimensions to a fairly great precision.
At the same time it may actually cost you real utility.
True enough in theory, but does it in practice. What cartridge gives me significant increases in utility that is excluded by my parameters? If the utility (value) is high enough, I might take on some additional risk. The reason for asking the question may be partly due to having already determined that the differences in value did not produce a clear winner. If there was an ideal cartridge in this space, like one with a 6mm bullet range from 40g to 140g and MVs of 4000 to 3000ft/s respectively, the question might be different, like "anyone want to buy a couple dozen .308 rifles, reloading dies, etc?"
An example is cases that are within resizing or trimming/processing range of each other. Choosing pairings that way (which will naturally give you cartridges that are in the same family and therefore break your "safety" rule)
I'm not sure this does break the "rule." The .243 Win is exactly the same case length as the .308 Win. Even though it locks in a .308, it looks like it could be benign, at first glance. It fails, but it most likely will fail safe. Poking around the internet a bit more I found this article:
http://www.levergun.com/articles/wrong.htm
The .243 Win is looking more attractive. It is definitely popular It looks to be quite flexible, low cost, etc.
Other benefits of similarity (magazine compatibility, storage ease, etc) are also lost.
My .308 mags for the CZ-550 are actually .243 mags. Could be a plus in going .243 then, or a minus... I have not seen an issue with sidearm calibers. I have at least ten magazines for every sidearm I own (each numbered to track them). Each sidearm caliber has its own distinctive reloading dies, plates, holders, boxes, shelf space, color, log book, storage bins, ammo cans, etc. I do the standard thing and put everything away between calibers when reloading. I have organized things to make this as painless as I know how. Only the consumable stock (powder, primers, and bullets) are located in a cartridge agnostic way. If I do a .243 CZ-550, I suspect I will be doing a bake on finish of the magazines dedicated for it to the new color for the .243 Win to distinguish them from my existing .308 ones, if I follow my SOP. For instance, currently my .45ACP mags go in a lockable container labeled and colored to the .45ACP color (blue). All .45ACP ammo is likewise in blue MTM boxes and/or in ammo cans labeled in that color, either as a paint spot or sticker. The reloading dies have a blue color spot on their mounting plate or are in a box with a blue sticker to identify them as well. A bin with a blue sticker and label holds many tools, such as dedicated .45ACP OAL gauges, reference weights, hole and pocket reamers, go/no-go gauges, etc. Another bin with a blue dot and label holds extra barrels, springs, and other small parts. Another bin with a blue dot and label holds the ziplock bags with fired brass. I think you have the idea - reuse is not a bit thing for me; I am more interested in neatness and correctness.
Okay -- enough with the Garanimals jokes...