Ok this one is a pain in the arse. The unblanking is taking too long to happen.
I've got two scope traces below.
TP588 (yellow) label is the output of the trigger generator.
BLK (light blue) is the z-axis output. As this increases it increases the intensity of the trace.
SWP (dark blue) is the horizontal sweep. As this decreases, the trace moves from left to right.
Q574 (pink) is the collector of the input transistor of the blanking circuit triggered by TP588.
The output at TP588 and Q574 are absolutely spot on and nice and sharp. BLK is however soggy and looks like it has an RC-time-constant function on it.
As you can see with a relatively slow trace, when the trace starts, there is enough time for the unblank to happen before the sweep gets very far.
On a relatively fast trace, it hasn't even finished unblanking before the trace is done
Aaaand I'm done for the day