A reproduceable kernel panic can sometimes at a stretch be considered a denial of service exploit, but in this context I want to see unsigned code execution.
there is no details but assuming there is controllable stack pointer, I would say it's just matter of time. taking into consideration how soon this device is in hands of researches I would be surprised if there wouldn't be dozens of exploitable bugs there.
given that we could count time to first crash in hours (it was just few days since people started getting those devices) we should expect code execution soon