I know this won't set the story straight forever but if you do an anal amount of research like I did into this issue for an as-yet unpublished article, including reading old, original C-22 design statements made by Frank Butler and reviewing the design with two experienced engineers, one of them the Dean of Engineering at a prestigious university and the other a 35-year veteran of GM, and perform all the calculations, you'll come to two unmistakable conclusions:
1. The bolt was NEVER intended to perform any function at all in a turtle incident. It was only intended to slow the keel down after it was knocked up by hitting an underwater object so that it would not damage the trunk on impact and potentially sink the boat. It has happened in at least one documented incident that I'm aware of and that's enough to scare any mfr. into installing some kind of safety device.
2. It could only possibly slow down but not stop the keel as it swung shut in a turtle incident under PERFECT conditions: maximum hand-applied torque, lock bolt in perfect condition with lubricant, perfect keel condition and coatings, no pivot assembly wear at all, very slow inversion, and so on. In the real world, it would not affect the outcome much, if at all.
I lock mine when I remember to. I've hit unexpected rocks once and had a near miss from hitting a mysterious piling in the middle of a bay during low water level. It could happen to you. You could get very wet.