I assume that the bit you have Michael is just the probing part (i.e. the switch)? So you won’t have to build up their frequency-hopping (FH) receiver to interface to your machine!
An interesting point about the system they describe in the patent is that it has to operate in a very noisy (in the EM sense) environment, using FH to minimise interference from other systems (WiFi etc) and also a clever protocol with acknowledgements and repetition to correct errors which introduces a fixed delay of the order of 5ms. As a result the control system has to allow and correct for this delay. When probing you move the tool towards the work at a low speed, and stop it when the probe is triggered. If there’s a delay before the trigger gets through it will move a bit too far, but you can back-compute the actual trigger point knowing the delay and movement speed. Of course the probe has to tolerate some over-travel so it doesn’t get broken! If you want to digitise a few thousand positions on a workpiece obviously you want to move the probe as quickly as possible
Mach 3 and Mach 4 actually start to decelerate the axis when the probe signal is detected and latch the probe position when the movement stops, so both give an inherent error. It seems that UCCNC controllers do a better job, they latch the position when the probe signal arrives and then decelerate.