Sticking an MV in an HPS or pulse start electronic ballast will destroy the start probes, you will notice that one end of the MV will get quickly sputtered, if you use a clear MV. A probe start MH ballast should be ok, althought there is still a 750V pulse but it is not that bad.
No, it won't destroy the probes. You may use whatever high voltage ignitor, but the voltage would never be above what the lamp allows, so in case of a probe start MV it is about 200V or so. And no, the resistor isn't at danger either, the same thing: The discharge will move to the main electrodes instantly and limit the voltage.
The voltage is always dictated by the lamp, so when the lamp wants 200V, there wont be more than 200V
The voltage printed on the ignitors is, how high they may go only when unrestricted by the lamp. Normally the lamp is supposed to restrict it below that voltage (by the gas breakdown). In fact if the lamp wont restrict it, it means the ignitor wont be able to start the lamp in the first place.
So no, starting cold lamps with a HV ignitor does not do any harm to them, nor the ignitor.
But only when the lamp is cold.
The problem with HV ignitors and probe start lamps is elsewhere:
When the lamp is hot, the breakdown voltage of the gas within the arctube is way higher. That is, why it needs to first cool down before it is able to ignite again. Because the breakdown voltage is high, so will be the voltage the lamp will limit to when e.g. getting pulse form the ignitor. And this may be way above what the internal construction in the outer or socket can withstand without risking its damage. The thing is, lamps that explicitely need HV for the normal ignition have to be designed so they withstand the maximum voltage the ignitor generates when the gas won't break down (e.g. because of the high pressure when hot,...).
But lamps designed to be started without an ignitor normally have to withstand just the couple 100's V the ballast OCV, so they are using different design of the outer (clearances between conductors, gas fill pressure and composition,...). And that construction may break down when exposed to the kV of the ignitor and e.g. start an arc in the outer. Don't forget the elevated pressure is in the arctube, not in the outer bulb. So when the arctube may have safely lower breakdown voltage than the outer when cold, the relation may reverse when the lamp was just extinguished (e.g. by power dip,...)
In electronic fluorescent ballasts the output current is limited by the heater cathodes and the low pressure gas, and usually the circuit is closed by a capacitor, but a HPMV is a short circuit, so it is more tricky to operate than a fluorescent tube.
No. Output current in fluorescent circuit is not limited by the electrodes, nor by the low pressure gas nor anything within the tube. Electrically a thin 6mA CCFL would easily carry many amps of current, only till its destruction of course. Which may happen pretty fast, but since then there is practically nothing to limit the current at all within the lamp itself.
So even with fluorescent circuits, the only thing that limits the current to whatever the lamp is designed for, is the ballast. Regardless what its construction is. So the current limiting component could be an inductor (series choke preheat ballasts, but as well as practically all of the mains electronic ballasts), it could be a capacitor (many CCFL ballasts, some night lights, many "royer oscillator" based DC input fluorescent ballasts like in emergency fixtures,...), it may be an output characteristic of a flyback inverter (the low cost battery powered inverters, some emergency fixtures,...), sometimes a resistor (fluorescent lights in old trams, old HV battery powered fluorescent lanterns, some early cheepeese light fixtures using incandescent lamps or even a resistor element,...), but it is always the ballast, what says what current will flow there.
The problem with high pressure lamps on electronic fluorescent ballasts is, fluorescent ballasts use to operate at high frequency AC. That is good for the low pressure style discharge, because it pushes the current density more towards the discharge surface, so operates it at lower current density where the light generation is more efficient and closer to the walls, so with less path for the light to travel so lower self absorbtion here. Plus high frequency means the reactive components (mainly coils) are smaller, so could easily be designed with way lower losses (mains frequency coil uses to dissipate about 5% of he VAs the coil is exposed at, the high frequency coil below 0.1% of he similar VAs; but with electronic ballasts the rest of the circuit means other losses).
But the high pressure lamps are usually small, so their acoustic resonance uses to be higher frequency close where the electronic ballasts operate and with high quality factor, so may develop very strong resonance effects leading to standing waves causing local overload of the arctube. There many ways to address this in HID ballasts (frequency wobbling, use of low frequency square wave feed,...), but as the fluorescent tubes do not exhibit these issues, ballasts designed for fluorescent feature no explicit measures against the lamp resonance.
