That did explain it much better... but IMO it still is not a good idea. I mean, while I clearly understand how it would help you being able to set provinces to no repair to better achieve your aims, I mean that the idea suffers problems of history, logic and complexity.
Not really as it denies the commander the ability to choose which areas to upgrade. To tell him that he must upgrade all provinces, including those of minor importance or which he may have no intention of holding does defy logic. Also I don't think it greatly adds to the complicity, at least not for the player. Don't know enough about the actual programming to say on that but if it is difficult that is a valid point.
That is intentional as "a brake on conquest speed" and IIRC was modified for greater difficulty some years ago. Think how many hours it takes to run your MOTs thru a province. Now compare to reality. If you can not repair at 100%, you are probably conquering too fast for game balance and the repair bill should be doubled again to get greater realism into the game.
Given that you have sometimes mentioned how quickly you have conquered assorted powers, markedly faster than in RL that does sound rather a strange argument.

I'm actually a fairly inexperienced player, largely due to lack of time. Don't think my attacks compare to the historical ones so if you do what you propose you might make it a more challenging game for the experienced player but even more unrealistic for players like me.
Also given how much the game diverges from reality, for instance massed British strategic bombing forces in 39-40, fairly easy [for many players] Sealion and very deep conquests in Russia, massed conquests and annexation by powers such as the US, very strange naval and supply rules, the bias in favour of continental powers in terms of manpower for instance, how over-powered paras are etc, then I find it strange that a strategic commander can't priorities repairs to any realistic degree. Note here even when you priorities a number of provinces that doesn't allow you to select say only transport repairs or give any idea what the actual total cost for the province is. [This latter is what I was mainly arguing again].
While you, I and other players may do that and it does indeed make for magnificent game play, IRL nobody dared do that without insuring that the path from supply depot to the panzers was repaired enough so the trucks could get to a new forward supply dump. In short, "deep penetration" in AoD means running 3-4 provinces or 1000+ km behind enemy lines while in real warfare it means going less than what tank can travel in one day maximum. And that was not even 100 km. Or it was out of gas. In short, deep penetration ends a few kilometers past where your fuel trucks can reach. But - with rapid prioritized repairs - true deep penetration is a reality.
I think the best argument to leave as is (or actually increase repair bills) is to handicap skilled players.
Actually I think your wrong here because it is the mobility of the forces with
unrepaired provinces when the enemy has largely collapsed that enables such deep penetration. Hence increasing the repair bills or making them more random would have no impact here. I'm talking here mainly using my experience of playing the Soviets and the sort of collapse of the German defences that occurs in 43-44. [Also started one game as the Germans but that has been paused for nearly a year because of other matters taking up my time]. If we wanted a truly realistic campaign we would need steps to prevent this, or thing such as the German advance to the Urals within months of attacking the Soviets but increasing the bill for already unrepaired provinces is unlikely to be a factor here.
Possibly what you would need to prevent this, which is what as you say occurred historically, is that units carry much less supply so they need to resupply far more frequently. Especially for motorised/armoured units. Basically units can make the 4-6 province penetration we're talking about on their integral supply without any addition needed.
Steve