Looking for opinions

No problem! Always happy to share.

I guess it matters what you mean by "falling off". If you physically just drop the tip a very small distance like others have mentioned here, inspecting the tip for damage is probably fine. If the program itself causes the wrist to rotate the tip into anything and it falls off, that I consider a crash. Both scenarios I would prefer not to take the chance and run a verification and make a judgment call. I'm personally responsible for my machines accuracy and I don't want to risk anything. Machinists already have a hard time trusting the CMMs. If they saw my machine crashing with no investigation, trust would be few and far between. Everything I do is for my customer. The machinists and the external customer. I'm sure everything I do is overkill, and I don't expect others to follow my lead, but if I hear anyone complain about machine to machine correlation or machine accuracy, they have no leg to stand on.

I usually don't share programs I write because I wrote it on the clock so it's now owned by my employer. That's why I gave a detailed description of what I do. I can definitely help you create one if you need the help though!

My last job used a similar setup. We would check all our probes at A0B0 an then at A90 for B0, B90, B180 & B270. The check was done at the beginning of each shift and if any probes reported out we'd calibrate all tips for that probe. After a collision we'd run the check and calibrate as necessary