I might take some pics of the home lab, work lab would be out the question as I love my job too much!
As for the mice..... When we have the need then its neck and glass rod every single time, not once in well over 30 years has it gone wrong. I am not a fan of CO2, then again looking back over my 30 odd years I would say we kill less than 0.1% of the numbers of mice that we used to. Hell most of the time everything is now modeled on a computer!
I think its a good thing that we have developed techniques that often make using mice etc a thing of the past, but even now on occasion there is a need to use a live model. I like the glass rod and pull method as its quick, and for the record you dont pull the tail you pull right at the very base of the tail so the spine is in a straight line.
So the full process is
Grab mouse by base of tail and lift from cage (or trap), have glass rod in other hand ready and as you bring the mouse to the tile it will stretch as it reaches to land on the tile. The moment the front paws touch the tile bring the rod horizontal and directly behind the neck and skull, firm but not so hard as to restrict any of the airway you push down (mainly as a stop point) and at the same time pull with the other hand holding the base of the tail.
Complete dislocation and severing of the spinal cord, sorry if it's hard to understand but I can assure you its very easy to show in person and quick to learn.
In the past I have worked with other larger animals and these are normally 'vet' assisted. Only bad experience I have seen or had was an escaped primate in the late 80's, very few options when something like this goes wrong and the animal is infected with a human transmissible pathogen. I havnt worked in this area for a long time now and so I have no idea what systems are in place, I would think safety and accident protocols are very much better than when I worked in the field.