A couple of very interesting issues, for sure.
Ricochets
Ricochets are some interesting counter-intuitive buggers. You never know what's going to happen until you test it out and even for projectile/target combos demonstrated "safe," it is the 1/1000 exception that bites you.
Steel plates/falling steel targets, for instance, give me the willies. Oh, I love nailing them in a course of fire, but I learned a whole lot this one time I went to a range that had wooden dividers between plate stations. All-lead projectiles looked to be pretty harmless if you kept velocity up. Jacketed projectiles were a whole lot less predictable and I saw jacket bits or entire jackets slung back to embed into the wood partitions a good ways back. This was for falling plates. I suspect plates that rotate on an axis or are rigid would be more dangerous.
I have shot under circumstances where the backstop is a highly or moderately ricochet-prone material. I did this while under cover with the exception of some of my head, shoulders, & arms necessary to send rounds downrange. I also had a k-pot, body armor, and issue go-fasters (glasses). I was also being paid to do so. Given the totality of the circumstances, I was not particularly concerned with ricochets.
Penetration Testing
I think this sort of activity is of value. Knowing how one's chosen SD/HD weapon/ammo combo will perform on materials in one's home or on commonly encountered materials outside the home is helpful in determining which weapon/ammo combo is best for you.
For instance, I am partial to slugs for SD/HD. I like them because they can be placed as a rifle/carbine round, hit like a freight train, are effective at range, and come in a package (my Rem 1100) I point & shoot faster than any rifle/carbine. But, my HD shotgun has all 000-buck in the mag tube because, although my house is brick and my neighborhood's houses are brick, the neighbor on the other sie of my front door has an addition to his house that is vinyl siding. That is their TV room and where they spend their evenings, many times. Any scenario that has goblins coming in the front door has mere vinyl, drywall, and distance as my backstop. A slug would penetrate my neighbors addition, but the distance is great enough that 000-buck would bounce off, most likely.
Daniel:
I think your experiments of value, but I do worry about your safety. I believe you when you write about the ricochet effects and probability. But, collisions between projectiles & materials are devilishly complex physics phenomena and there is a low probability that you are going to eat a ricochet even with something like 00 buck vs brick. That probability exists with each round fired.
I think there are some precautions you can take that are relatively painless that will drive down the probability of injury due to ricochet.
1. Eye protection. Some glasses/goggles with large lenses can protect the bits most vulnerable to injury due to ricochet.
2. Head protection. A k-pot is likely overkill, as is a milsurp steel pot. Though, a steel pot has a "cool" factor in my book. Milsurp steel pots can be had reasonably and can be cocked down to cover your forehead and a good portion of your eye (down to the top of the pupil).
A less cool option is a construction hard hat found at Home Depot or Lowes. As above, adjust so that it covers your forehead & top part of eye.
3. Body protection. No, I don't suggest body armor. I think that a small sheet of plywood, with two legs arranged how you feel best, and a cutout padded to act as a gun rest, would do the trick. The gun rest would aid in shot placement, as an added bonus.
I would also suggest the plywood be placed at an angle, so the effective thickness vs a projectile would be greater. Last, a small (inches) lip orthogonal to the plywood placed below the gun rest to catch any ricochet that wants to ride the plywood up.
Good luck with your testing and stay safe.