These molar macs were known for the snapping even when in service. Back when I was in high school, I was the IT assistant for the school, this was in 2003/2004. Right at the tail-end of these machines service life, and were being replaced by the dell 2400 series black/gray towers in droves. They were hot shit back then.
And the last few remaining were starting to snap. I ended up figuring out that it was 2 different things. on one, it turned out to be the HV regulator caps that keep the high voltage stable were starting to fail and causing the HV to drift way high.
The second one actually was the Focus/G2 block embedded in the flyback that started to fail, and it would intermittently toss 23Kv into the CRT neck focus/screen pins, which are only rated for a KV or two maximum at best.
Then you got the stupid cases where dust/dirt makes it into the anode cap and arcs through it to ground. Then you start to smell corona at that point. What you wanna look for are the spark-gaps on the CRT neck board, when the snap occurs, if you see arcs jump through the sparkgaps on the neck PCB, its because the flyback is shorting on the G2/Focus divider.