To prove the point... Try an initial torque of 60 and an accel of 5
Fail
You choosed too extreme values, the LSD act strangely with these since GT2.
Try to stay in 8-50 range to correctlly understand initial : you'll see that the main effect is LSD's "strongness".
If you go in 7 or under, or 50 or higher (I beleive the values are % of traction power to the spinning wheel but I won't say I'm sure), you'll profit of what I call "the dark school" of the LSD : these are opened or closed values, and makes the exact effects you describe, shutting the LSD's main effect. At this values the "dark initial" acts like "normal accel/decel" and "dark accel/decel" acts like "normal initial".
But in the "normal" range you'll see initial is the LSD strength.
To see that, there's something more simple.
Have a very powerfull and very ligth MR, put accel to 49. On a straigth line, try to stay with one traction wheel in the grass and the other traction wheel on the road @150km/h.
With initial @ 10, you'll autoturn back to the road or the car remains semi-stable in grass.
With initial @ 30, the car will want to stay in the grass.
With initial @ 49, you'll be doing donuts @150km/h and crash.
At 5 you have a dark initial (open LSD) so the LSD won't even lock so nothing can be seen and if it locks, it locks with a strength of 5 so who cares
At 60, I'm not sure but I think it become too much sensitive because of the "dark effect" and the strength of 60 is too sensitive (initial becoming a accel/deccel effect). I never use closed LSD on MR anyway, it's dumb and noobish lsd work even for a drifting setup.
But how can't initial be the LSD strength, then, within normal values. It's just his main effect.
What you describe is lsd's dark school, something you see only when you're setting your car a very special way to seperate deceleration strength and accel strength using a locked sensitivity with initial <8 or >55.
This method is what you want for exemple on the yellowbird to correct understeer to oversteer to understeer problems. Just try and tell me.
This effect is the main reason nobody understand lsd correctly.
My personal theory is if these are really %, well at more than 50% you're stealing unspinning traction power to give the same to the spinning wheel and the LSD never stabilize because of wheel inertia. And <7 I just don't know. Maybe the same because of wheel inertia.
I allready made a wall of text on a topic with guys asking how a lsd worked describing that...
What do you think ?
@OT : your default 4RW is not good unless you have a 50/50 traction car with a 50/50 mass (MR engine position, like the ZZII, which mass distrib is 40% front/60% rear), and as every single one guy here you keep inverting dampers. Please keep on, the differences are "small" but we'll see who will win tuning comp