Profile of yuxin

Disclaimer:
This is not a I-am-smarter-than-the-devs-and-they-should-implement-this thread. I am a Computer Science student with a concentration in game development. I am interested in game design and mechanics implementation. I'm posting this here to get feedback on an idea I came up with last night.

Overview:
Manual vehicle pathing would involve players defining a path for the vehicle by manually drawing a path on the terrain for the vehicle to follow. Instead of doing any pathfinding, the vehicle AI would only handle steering and arrival. There should be some verification and smoothing/lerping of the path.
This control scheme is more popular in touch-based mobile games but I think it has great potential in RTS games.

Image from "Flight Control" , iOS mobile game.

Story:
I was playing World of Tanks with a friend last night. I was in a KV-1S and a fight happened at a choke point. A friendly scout tank (Luchs) was blocking the choke while I really needed to pass to kill an enemy heavy tank (IS) that the scout tank had no chance of penetrating. So I pushed the scout out of the way and got the job done.
I realized that a similar engagement in CoH2 could have been a dead scout car and a heavy tank with rear pointed at the enemy if not micro'ed carefully.

What problem does manual pathing solve in CoH2?
Tanks "derping out".
"Scheisse, zat mortar squad is in the ze way. Turn around and go in front of ze AT gun!!"

Is it the same as shift queuing move commands?
No. In between each shift queued command, the vehicle AI would still attempt to pathfind and if there happen to be something blocking the way, the AI may "derp out" and go all the way around the map.

Can similar pathing issues be solved by other implementations?
Yes. Sudden Strike series had a priority list of sorts that more mobile vehicles will get out of the way of the less mobile ones. The choke point blocking vehicles will get out of the way if the other way around is far. I don't know how they did it but I think it's something along the lines of this heuristic
Cost = distance to target * my vehicle's fatness + SUM(tell vehicle to move out of the way distance * that vehicle's fatness).

Other (insignificant) advantages of manual pathing:
1. Computationally cheaper than path find every command (think in case of spam clicking or shift queuing).
2. Players shouldn't use "oh the pathing fucked up" as excuse for losing vehicles. The fewer losses players can blame the game instead of themselves, the more competitive the game is.

Final thoughts:
Manual pathing should never take place of automatic pathfinding. It should only be an option to micromanage tactical movements. With manually defined path, the vehicle can either ram/push/plow through obstacles if they are capable of doing so or wait on the path until the path is cleared.

Criticisms are very welcome.

tl;dr : Player drawing a path for vehicles to plow through.

I'd personally give them 40% engine damage, 10-15% chance to immobilize and increase their cost. The soviets don't use nearly as much Munis as the Axis.

Wouldn't that 10-15% on a heavy vehicle become too much of a gamble when the cost is increased?