Hi Fruitbat - this seems like an odd set of symptoms, but there may be an explanation. First, its definately a 'Sud rev counter? Certainly the early ones, like my '78 Ti used discrete components (not IC's), which may differ from later Sprints and 33's.
The principle of operation is much the same; the pulse train from the points is integrated to produce a voltage proportional to the engine speed (frequency-to-voltage conversion) and then this level is used in turn to drive a sensitive milliammeter (the rev counter needle).
On early rev counters (maybe late models too), the integration is performed by a capacitor/resistor network with some transistors for amplification/meter drive. If the pulse train from the points is noisy, what should be a well-defined train of pulses may mush into a constant voltage at high engine speeds, and this would cause the rev counter to drop to zero.
On the face of it, this may explain your problem. I would check your points (if you have them), condenser (ditto) and HT leads first. Make sure your coil is not breaking down internally, and is connected the right way round (+Ve to ignition supply, -Ve to points/electronic ignition module and rev counter sense wire on this terminal or dedicated module output) and whether the sense wire is routed near any HT leads.
Also check the distributor cap for tracking and maybe the ignition timing if there is a possibility that "spark scatter" may be occurring.
If all that checks out, there may be a problem with the rev counter itself, but would be more likely to be the meter movement rather than the electronics, which would probably fail completely (as mine did
) rather than act intermittently.