Just a detail about cooling the brass, is the quench in water necessary? From what I have read about cartridge annealing air cooling in ambient air should be sufficient. Water cooling doesn't hurt but it adds a drying step to the whole process.
There are two good reasons to quench.
You want to prevent the body of the case from annealing.
You want to stop the annealing.
Annealing the case body has issues, but more importantly, annealing the case head can cause major problems even on bolt rifles.
To anneal brass, the temp should be held at ~650 F for about 2 minutes. Since this would result in the heat spreading down the case, the more common practice is to heat to about 750 F for a few seconds. Since you do not want to fully anneal the case neck, the annealing must be stopped - so quenching is the common practice. If the case neck is over-annealed, it will not grip the bullet securely. Repeatability is fairly important in annealing, especially if the bullet is not crimped (a common practice in precision shooting).
I have always had some doubts that these annealers actually produce good repeatability when RPM and dwell time is blindly followed.
The roller will heat as the machine is being run, so it will be actively cooling the first cases, and actively heating the last cases. This will have an effect on the rate at which the case necks reach the proper temperature. One would really need to have a bunch of old cases that would be run through the machine to get it into equilibrium. It is easier to just watch it run and adjust it on the fly. The flame temp will also vary depending on the temp of the propane bottle and amount of propane remaining.
I think the platter style annealers answer most of the problems better than the vertical style.
One idea I had was to use a temp sensor and an Arduino to eject the case at the proper temp.