Something to be aware-of is the elevated current used in multiplexing, in this case 10mA vs nominal 2.5mA. Info I've read says nixie-degradation is an exponential function of drive-current (somewhere I saw 2.5).
Even if the exponent is 2.0, increasing the current 4X and multiplexing at 50% would cause tubes to degrade about 8 times faster than they would with direct-drive. For this reason, and others, I always use direct-drive for discrete tubes. The one panaplex design I did had separate pins for all anodes & cathodes, and that, too, was direct-drive. (most multi-digit, and probably all dot-matrix displays require multiplexing due to pinout).
In this particular case, where only 1 tube failed, it's more likely to be a manufacturing defect, or the tube had substantially more hours of usage than the others.