I don't know if clutch pack style is better for track.
I do know that they are best for drag racing as a clutch pack behaves as a solid axle when power is applied (more power=more lockup force). So there is no rpm difference possible as long as power is applied. Great for a straight line as the drivetrain has an automatic tendency for that once throttle is applied. Thats why they're popular in the US for an aftermarket install. The land of the 1/4mile.
M diff is not clutch pack style but gerotor actuated clutch pack. It does not lock under power but under rpm difference. It has greater lock up force possible than the atb or clutch pack (I think), but lockup is created by rpm difference, not by torque needed. So there needs to be loss of traction before lockup can occur. (or lockup can be induced in tight corners even if no traction is lost)
For the /m type gerotor diff, 10rpm difference on both wheels is (for example) 200Nm lockup force, 20 rpm difference is 400Nm lockup force etc
Also hysteresis is a risk (pulsating on/off; lockup induces 0rpm difference->lockup is lost->rpm difference->lockup->lockup induces 0rpm difference->lockup is lost->etc). This pulsating is compensated by dampening, but dampening the gerotor also means that lockup and freeing up the diff gets slower.
So they act on 3 totally diffent factors:
clutch pack: lockup on power (more power=more lockup)
gerotor/m diff: lockup on rpm difference (more rpm difference, more lockup)
ATB/quaife lsd: torque biasing: the wheel that is unable to convert the torque to traction induces lockup. That means if the wheel is unable to roll along with the car (inner wheel wants to spin, outer wheel doesnt accelerate accordingly), lockup is induced. In theory rpm difference is no issue as it is a balancing system, but lots of rpm difference means more friction on the cogs, so that is a force that may help induce lockup partially (I don't know how much this is an issue). Also lockup is in a ratio of torque (e.g. 1:4)