I can't comment on Pawpaw's design, he used Solidworks to model the flow dynamics and seemed to know what he was doing. I do agree that a compressed air is better and that's what I use on my CO2.
It depends which firmware you use, either D9 for both N.O. in parallel or you can do D9 for X and D10 for Y and use N.O. or N.C. switches.
On my laser I use D9 and parallel switches but on my larger CNC I prefer to use N.C. switches for safety. If a wire or switch fails it will stop instead of driving into the end stop and potentially damaging a ball screw etc.