Last year a friend had a G64 that fired the ejection charge right after ignition. I asked if he remembered whether the delay insulator was loose in the well or not. He did remember that it was very loose. I told him about the tape method to tighten the delay insulator up. I ran into some very loose insulators also but thanks to a thread on here a couple years ago about the situation I learned to use the tape method. Never had a problem. I know Handeman posted to that thread as well as myself will try to "dig" it up.
Found it/ one of them, man how time flies:
https://www.rocketryforum.com/threads/at-29-40-120-early-deployment-issue.136925/#post-1632289
One would think that the O-ring would seal the gasses with a loose fitting delay insulator but apparently it is not.