I have two Growers Supply Digital 400W ballasts that can operate HPS,ProbeMH,and PSMH.Just put in the bulb-the ballast adjust automacally.These do cause RFI-so I stopped using them.Besides the magnetic ballast works better for its particular lamp.

So this magnetic ballast likely just changes the taps, without removing the ignitor.

When there is a mechanic switch to select the lamp type, I would even expect this switch to deactivate the ignitor as well (or at least I would wire it that way, it cost the same).
Normally, when the lamp is cold, even with the ignitor present, the main arctube limits the voltage so, the discharge in the outer can not happen at all. It is only in the case of hot restrike, when there could be problems.
The electronic may contain an identification of the arc in the outer (voltage profile, noise) and so be able to shut down for some time to first extinguish it and then let the lamp cool down (so make the breakdown voltage within the arctube lower than in the outer) before the next reignition attempt. As the discharge in the outer could happen only during a hot restart attempt, it is quite rare and when the ballast responds correctly only of very short duration, so the lamp assembly will most likely survive that without any significant damage.
It may be even programmed so, it tries to first ignite the lamp without the use of the HV pulses (so when there is a probe MH, it will already start) and apply the pulses only after some time. And to protect against hot restrike attempts, it may easily contain a timer (powered from a capacitor backup, so it counts even without the mains), which will prevent HV pulse application for a predefined time (half hour or so) after it detect the arc has extinguished (by itself or by power interruption or so). With that it would be able to prevent application of the HV pulses on a good probe start lamp without any reconfiguration.

Technically would be way better to disable the ignitor by some means of configuration (switch, jumper wire,...; so no confused automatic may destroy the lamp) but there is problem with many users, who do not have any clue about HID's and so configuring it wrong as the result. So for practical use the electronic, which is able to automatically select the lamp, is the best - no one could ever configure it in any wrong way...
And the automatic selection is practical even when you have many fixtures with different plants (so different color needs) and all the ballast remotely in some electrical cabinet: There is quite likely an error, where you switch over the wrong ballast (belonging to other lamp). The automatic operation just operates the lamps according to what it detects, so most likely according to what really is connected to each of the ballasts.

When there is a mechanic switch to select the lamp type, I would even expect this switch to deactivate the ignitor as well (or at least I would wire it that way, it cost the same).
Normally, when the lamp is cold, even with the ignitor present, the main arctube limits the voltage so, the discharge in the outer can not happen at all. It is only in the case of hot restrike, when there could be problems.
The electronic may contain an identification of the arc in the outer (voltage profile, noise) and so be able to shut down for some time to first extinguish it and then let the lamp cool down (so make the breakdown voltage within the arctube lower than in the outer) before the next reignition attempt. As the discharge in the outer could happen only during a hot restart attempt, it is quite rare and when the ballast responds correctly only of very short duration, so the lamp assembly will most likely survive that without any significant damage.
It may be even programmed so, it tries to first ignite the lamp without the use of the HV pulses (so when there is a probe MH, it will already start) and apply the pulses only after some time. And to protect against hot restrike attempts, it may easily contain a timer (powered from a capacitor backup, so it counts even without the mains), which will prevent HV pulse application for a predefined time (half hour or so) after it detect the arc has extinguished (by itself or by power interruption or so). With that it would be able to prevent application of the HV pulses on a good probe start lamp without any reconfiguration.

Technically would be way better to disable the ignitor by some means of configuration (switch, jumper wire,...; so no confused automatic may destroy the lamp) but there is problem with many users, who do not have any clue about HID's and so configuring it wrong as the result. So for practical use the electronic, which is able to automatically select the lamp, is the best - no one could ever configure it in any wrong way...
And the automatic selection is practical even when you have many fixtures with different plants (so different color needs) and all the ballast remotely in some electrical cabinet: There is quite likely an error, where you switch over the wrong ballast (belonging to other lamp). The automatic operation just operates the lamps according to what it detects, so most likely according to what really is connected to each of the ballasts.

I would trust the ingitor being off more than an automatic ballast deciding for me.
And your reasoning is likely why probe start MH bulbs work on my electronic ballast.

I have no experience with probe start MH, but I do have experience from experimenting with the probe start MV on fully equipped HPS ballast (talking about 230V area, so both are just a series reactors on a 230VAC input):
I've never seen the ignitor to cause any arc in the outer (however it can not be excluded, unless the lamps are designed for it), the only observation I made was the MV ability to instant start when there was the ignitor.
Because I don't think the probe MH's are so much different from the MV's, I may conclude the arc in the outer as the result of the ignitor HV pulses is rather rare thing, so in most cases you won't observe any issues. It could have been just a side effect of the actual lamp design, or just not enough "luck" to meet a marginal lamp, where it would be a problem, the lamps were rather fresh (nearly no wear), I hadn't do the experiments long enough to catch the chance,...
But the thing is, standard lamp specifications are not designed for that, even when it seems to work, no one will guarantee it won't trigger it  after a week of use, after the lamp ages a bit, after replacing the lamp,... The chance may be rather low, but the problem is, once it happens (and it is not treated properly by the ballast), it is 100% destructive for the lamp...

But I don't know, whether the lamps you are using are not really rated even for ballasts with the ignitor present (it is not that difficult to design even a probe start lamp to withstand the HV pulses from the ignitor, it just should be part of the design requirements for that lamp design; usually it is just matter for arranging the internals with sufficient minimum distances). If such ballasts are common for the horticulture, I would expect the lamps designed specifically for that use to be just made compatible with it, so in that case there won't be any problem at all...

But the odd thing is that all of these horticulture lamps seem to not be specified much for how to power them, as if the makers hardly know anything about proper lamp driving, and most of them seem to be probe start while being sold with pulse start ballasts.
But then it seems like the ballasts know if your lamp is probe start or pulse start (the electronic ballasts) and don't actually tell you anything about what strike type they use.
Or that the ballasts with these bulbs happen to be a lucky fir for not being destroyed.
My Nanolux ballast has not been tested with probe start lamps, but if I did, I would try some junk bulb, or my Ultra Sun one.

I have 2 125W MBF's that were used with Ignitors on 100W MH gear and were damaged by it

In one the starting resistor is burned out

In one the stem press of the outer lamp is melted off and wires broken, so the frame + arctube are laying loose in the outer lamp and scratch the phosphor. What i think happened is, the initial flashover started on the surface of the stem press, then it melted the glass, then the glass kept heating resistively by itself melting away and meling the wires

@MV damage at pulse start ballasts:
As I wrote, my experience was it survives, but I mentionned it does not have to be always the case. It could be, some makers make the voltage robustness lower due to some reason (don't expect any conspiracy there, it was just a side effect of some design feature and because the lamp does not have to withstand higher voltage, no one in the lamp development ever cared).


Horticulture lamp power ratings:
As I understood, the horticulture lamps are sold as specialty lamps, so intended to operate only on the dedicated ballasts and only one type (as per basic output characteristics) ballast is made for them, so the lamp vs ballast matching is clear.
The real rated power may vary among the different lamp types, even when all are intended for the same ballast.
And if that ballast is supposed to feature an ignitor, all the lamps will be designed to handle it well, regardless if they use starting probes or do not.

Generally there is no reason, why the probe start lamp should not withstand the HV pulses just because it is probe start.
The voltage limit comes from the construction inside of the outer bulb. Of course, a lamp designed to be ignited by a high voltage pulse must have the outer designed to withstand that voltage.
Standard probe start lamps are designed to work on just low OCV without the ignitor are not expected to be exposed to any high violtage, so makers just do not care. It does not mean it will never withstand that voltage, just no one cares about it. So it may or it may not, depend on the piece, it's state, but as well on the exact lamp design.
But if a lamp is supposed to work on a standardized ballast which does feature an ignitor, the design should always correspond to the HV requirements of that ignitor. The lamp may still need probes, because it is possible, the low energy ignition pulses are not enough with the given OCV (mainly to maintain the cold electrode discharge intense enough to actually warm up the electrodes)