I'll have to get it really hot again. After an hour or so it will do ~12V at 70mph, and it's easily noticeable in the headlights and dash lights. That's the part to me that really indicates trouble.32
Looks strange, yes, but it's not to me because I've seen it doing that since the first day I hooked a computer up to it like.... forever ago.
New question: what's the lowest normal operating voltage, at idle, that I should be observing?
Battery was tested on one of those big ass load testers at 300A for a seemingly long time (probably 6-10 seconds). Didn't drop below 11-ish volts at that draw. Started out mid-to-high 12's, vehicle off. . . .
Lower alternator/battery voltage is normal as the engine compartment and the alternator and battery warm up. It's needed, because at higher temperatures the battery can be damaged by higher voltage/current being pushed through it, and it's not necessary. However, I don't think dropping to 12 V, especially when cruising, is normal.
This brings me to the duty cycle signal, which might be the key to your issue, and the explanation of it's role in the Mitsubishi alternator (AC Generator Design Differences
As was explained in the article, when the voltage coming from the ECM to the base of TR1 is low, TR1 is off. The internal regulator samples the system voltage ("S") at the junction of R1 and R2. With TR1 off, the sample is 2/3 of "S". In the example, if "S" is 12 V, the sample is 8 V. At 8V, the zener diode will conduct, turning TR2 On, and that in turn switches TR3 off, cutting current to the field coil and shutting off the alternator charging output. (This is just an example; the actual point at which the zener conducts could well be different, but the impact of the duty cycle %, as expanded below, is the focus.)
However, when the duty cycle is at a high point, TR1 is turned on, shunting R3. Now, the sample voltage is only 1/2 of "S", i.e, lower, and the zener diode does not conduct. This, instead, causes TR3 to conduct turning on the field coil current and alternator output.
Consequently, the longer the duty cycle signal remains low, the less time there's field coil current and alternator output. Given that others have found the duty cycle signal to be significantly greater than "0", and if, in fact, the voltage at terminal 3 of the alternator connector is remaining low pretty well all the time, that would tend to hold the alternator output at a significantly lower average than when there is a 50% duty cycle with TR1 conducting 50% of the time.
That's why it would be helpful to confirm whether or not the voltage at terminal 3 of the alternator connector is in fact low all the time, in other words, verifying that the Romraider indication is correct. If it is, then I think we have to figure out why the duty cycle is remaining low.
The measurement can be done at idle, even before the engine is fully warmed up. Romraider shows the duty cycle at zero all the time, and I doubt the duty cycle should ever be at zero when the engine is running.
Incidentally, when the engine is running, two of the wires at the alternator connector should have system voltage; the third is the line from the ECM.