I have four way trim on a hat switch on each of my sticks - no fancy diodes or anything. I'm just trusting that only one person at a time will have a hand on a stick, and therefore only one person at a time will have a thumb on the hat. I guess it is possible that someone could bump the switch exactly when the other person is flying and trimming, but that's a heck of a coincidence. If your trim switch is mounted elsewhere, the question is why would someone else touch it? If it is mounted in a place easy to hit by accident, that might be a bad idea in the first place. A second pilot who is the pilot flying could bump it and not know it, if it is that easy to do so, and put the plane dangerously out of trim. With a selector switch, you could be unable to quickly correct the error, once it is identified. On the other hand, if it is somewhere that only the pilot flying will touch it, like my hat switch, the chances that the other person trims in the opposite direction are very low.
If you use a diode, imagine that the other person somehow hits the trim switch wherever you located it and runs it to full limit (a human induced runaway trim). Their action would trigger the diode and lock you out from running it in the other direction until they took pressure off the switch that they may not know they are pressing. If a CB trips when the switch is pressed in opposite directions, at least that cuts out the runaway trim before it goes all the way to the limit.
How many times do you trim during a flight? I set mine for takeoff during runup, then a few quick bursts at different cruise speeds (I just flick mine when it needs it - a quarter second bump), then a longer trim (half a second?) as I slow to pattern speeds. That's about it. Maybe 5-10 seconds total run time during a complete flight. Do you need to add complexity (and points of failure) in switches and diodes for 5-10 seconds of potential conflict?