I took a quick look at the circuit, and I think the Q1 spot should have a medium-high Vcbo transistor, like 120-150V, for best reliability. I don't know what the original part was. Under normal conditions or partial overload, Q1 doesn't need to drop much, but with the +75V (or the "+77V") supply shorted to ground, the bottom of the raw supply (C1's can), hence Q1's collector, will need to go to whatever the raw voltage is, maybe -100V or more. Such an event could have caused the C-E short mentioned in the OP, if the part couldn't handle it. I'd find it hard to believe that HP would not have used a suitable part for Q1, that could take a common fault like this. That's also why Q1 and Q2 are different - if Q1 didn't need to be higher voltage, they could likely have been the same part type. Alternatively, you could just say so what if Q1 breaks down when +75V is shorted. If it's stout enough, it will take a momentary short, or perhaps blow the line fuse, and not be damaged. I'd rather have it act more civilly, and just recover when the fault is removed.
The parts list says Q1 is silicon. It looks like the normal load on the +75V is around 25 mA, so modest power then. It looks like the circuit including CR703 and CR704 provides some foldback limiting, but I think the maximum would be around 50 mA, so worst case power dissipation maybe 6W, which should be no sweat for a properly chassis-mounted TO-3.
I'd recommend looking for a TO-3 Si PNP with Vcbo or Vceo over 150V, and DC-SOA rating better than say 100 mA at 150V.? I vaguely recall there are some medium-high voltage TO-3 PNPs used in the HP141T, that may be suitable, and very common, but don't recall the number (-0140- or -0138- seem to pop into my brain, but maybe entirely wrong).
Like Dave W. mentioned, you'd probably want something fairly slow, but simple circuit mods can fix any instability if choices are limited. Also, it doesn't necessarily have to be TO-3, since it's fairly easy to fit T0-220 and others into the footprint. I'd recommend not using ones intended specifically for SMPS use, without first checking the hfe spec at low collector current - sometimes they're quite low, which could cause circuit problems.
BTW my assessments are based on a manual I found online: 03400-90013, 1 December 1986. Others may show different circuits and changes.
Ed
