For it to be super-linear an additional meter of concrete / earth / whatever must be easier to penetrate than the one before it which I would classify as a physical impossibility. This is why linear is the ideal case.
Even if I were to accept the dubious premise that there is enough fractured rock to make a difference and there is no hampering with rocks falling into the void and that it's possible to hit the exact same spot repeatedly without touching the sides, all that would do in big O notation would be increase the constant factor. It would not be super linear after the second bomb.
If your are talking about bombs that hit side by side then clearly that is sub-linear as no matter how fractured the rock itβs not easier to push through than air.
An explosion creates a pressure wave. A pressure wave fractures rock. Fractured rock may be easier to pierce than solid rock.
Ergo, if first bunker buster penetrates to maximum depth -20m and then explodes, fracturing rock within a __ radius, then second bunker buster travels through that fractured rock, the second (and so on) may be able to penetrate deeper.
I have no idea about the physics of penetrating fractured vs non-fractured rock, but it's a physically plausible mechanism.
Furthermore, given the multi-minute timeline reported, there's enough time for the bombs to be deployed sequentially.