Scott, I believe that one of your friends is also a friend of mine
so I've had plenty of time to think about this issue.
I think that part of the problem here is that IR output isn't directly proportional to temperature. For instance, 800 deg. doesn't pump out twice as much IR as 400. The way I understand it, you basically have a temp where a material almost glows and that's some IR, but once you hit glowing, the IR goes through the roof. A 700ish Neapolitan oven setting has a cooler dome than an 850ish Neapolitan setting. That cooler dome is putting out a fraction of the IR. Without the IR, the hearth is using up it's stored heat from the conductive-rich initial firing (embers in direct contact), but not being replenished.
This doesn't help people with existing ovens, but I think thermal mass can help- to an extent. Your buddy with 4.5" walls- what I'd really like to know is the thickness of his hearth. A thicker hearth is going to store more heat, and thus lose less of it's overall energy with every pie baked. You'll still have the same deficit in IR replenishment, but... you'll bake a lot more pies before you start getting into trouble.
Bobino and Tampa Dave came up with a clad concept. A thin-ish firebrick floor with a layer of steel underneath. The steel would help to draw heat from the fire area, as well as help to even out the temp of the hearth overall, much like the advantage of an aluminum core in a stainless frying pan. If there's enough hearth real estate, I came up with the idea of targeted cladding. Splitting the fire area floor into half traditional firebrick and half steel with the same steel plate running under the rest of the floor. You start with the fire on the firebrick area, and, if the floor temp starts to drop, you move the fire to the steel area and feed heat faster to the floor. Into this mix, you could even add a third mode.
1. Fire on steel (fastest floor feed)
2. Fire on brick (slower)
3. Fire raised on grates (slowest)
If, say, the cladding ends up drawing too much heat for 850ish high temp bakes, you can raise the fire off the hearth and remove some of the conduction from the equation.
This is all extremely theoretical, and would involve a lot of testing to get the right thickness of materials, but the aspect that appeals to me the most is the amount of control.