Wayne,
Yes, the odd pulse appears at the 6AL5, and it's there whenever the flyback is in operation with signal at the antenna or not. What happens with a signal present is that odd signal is triggering the oscillator and the normal sync pulse that should be locking to horizontal just gets taken along for the ride so to speak.
It looks like this one with signal applied:
I don't see how that pulse could get coupled into the phase detector either, since there's no connection present. But that 6AL5 is positioned right below the flyback, so I suppose it's possible the pulse has such a high amplitude its radiating into the circuit and the 6AL5 is the antenna. There's no shield around it surprisingly enough, but I could try adding one to see if it helps matters.
*just noticed the normal sync pulse does indeed have a higher amplitude than the spurious one does, but once horizontal is locked to the spurious pulse there's apparently nothing I can do to get the circuit to 'choose' the correct sync pulse and lock to that. I have to get the spurious pulse out before this thing is going to work properly.*