I think the main point here is, that there are often some requirements to reduce the attack surface of a gateway. Based on what the topic-starter wrote, obviously even to reduce the attack surface when access to local console is already acquired (by physical access or breaking into LOM).
Something like: Access to local console should not allow to bypass access control mechanism e.g. by using GRUB shell to boot into single user mode (boots directly to root shell without authentication).
When you have a blackbox-like gateway from a vendor, you can only verify the verbal requirements against the actual behavior of the box or ask the vendor. If it does not comply and you cannot change it, you have to ask the vendor.
Because Check Point has some more whitebox-like approach with its gateways, which means you know that it is based on RHEL 7 these days and you even get a root shell without some special TAC script, you can check the implementation of some features yourself and even fix / change it. To ask vendor before doing so, is often a good idea when you want a supported setup.
I think this is was the topic-starter wanted to do here. But let's see, what he says.