I want to build some mini parking lot fixtures that use 50w MV lamps, but I obviously don't have any 50w MV ballasts, but I have 3 2 lamp f40t12 HPF ballasts. They run one 50w MV lamp fine, but would they run 2  lamps without overheating? I plan on putting 2 fixtures with one lamp on each pole, so they would have to be remote ballasted. That way, they wouldn't get overheated from the lamps. So would it work?

Also, if it does work, would an f48t12 HO ballast drive the lamps at higher power? I would like an answer soon, because it is going to be getting too cold to build these fixtures outdoors.
Thanks.

 Wont work too well.  Here why 50 watt 95 arc volts and .6 amp arc current.  While F40 ballast is run lamps at 110 volt at 430 mA.  So u got about 200 ma difference
As for HO ballast it good for 800 mA and a good overdrive in lamp current as well by 200 mA

Only real way is runing MV on 58 watt ballast or choke
  Or a MV gear.   U have better chance running 80
Watt MV on HO ballast.

Wont work too well.  Here why 50 watt 95 arc volts and .6 amp arc current.  While F40 ballast is run lamps at 110 volt at 430 mA.  So u got about 200 ma difference
As for HO ballast it good for 800 mA and a good overdrive in lamp current as well by 200 mA

Only real way is runing MV on 58 watt ballast or choke
  Or a MV gear.   U have better chance running 80
Watt MV on HO ballast.

I have seen members here run them on the exact same ballasts that I have. They have had no issues with them working, and they seemed to run the lamps at near full power. I can find you the links if you want.

But they never showed current the lamp is receiving if lamp run up all the way.  And. To run ballast on full terms.

If the lamps warm up correctly (till the unsaturated vapor state), the ballast would be fine - the arc voltage of the MV is almost the same as of the F40.
However the problem would be, if the lamp won't be able to warm up enough, then the arc voltage remains low and that may harm the ballast.
The lamp gets just about 36W and that does not do that much miracles, the MV's are already designed "on the edge" with the thermal budget at nominal power, so the risk of problems with that reduced power is not small.
Most prominent would be the position dependency - the best chances are with horizontal burning (gravity pulls the condensing mercury from behind the electrodes back into the middle of the tube, where the temperature is higher), worst would be vertical with starting probe on the bottom end (the condensed mercury may short out the probe and so prevent further ignition).

If the lamps warm up correctly (till the unsaturated vapor state), the ballast would be fine - the arc voltage of the MV is almost the same as of the F40.
However the problem would be, if the lamp won't be able to warm up enough, then the arc voltage remains low and that may harm the ballast.
The lamp gets just about 36W and that does not do that much miracles, the MV's are already designed "on the edge" with the thermal budget at nominal power, so the risk of problems with that reduced power is not small.
Most prominent would be the position dependency - the best chances are with horizontal burning (gravity pulls the condensing mercury from behind the electrodes back into the middle of the tube, where the temperature is higher), worst would be vertical with starting probe on the bottom end (the condensed mercury may short out the probe and so prevent further ignition).

I have tested 1 50w merc on the ballast in base-up to prevent the probe from shorting, and it warmed up to what seemed like full power. What would the normal arc voltage be? I measured it while running the lamp on an NPF Preheat f40 ballast, and it was around 90-100.

If it is around 100V, it is OK.
But chec, how the voltage evolves during the whole warmup.
The thing is, the problem is during the warming up, where the arc voltage is still lower, so the power delivered to the lamp is lower as well.  And in that mode the lamp may get "stuck" somewhere in the middle of the warmup - to see, if there aren't any stages, where the arc voltage would get "stuck" for some time or so, indicating the power is really marginal. Quite good indication would be, if you have the possibility to check the same at a bit reduced mains voltage (best is some transformer or so). The warmup profile will of course lengthen in time, but the question if the lengthening is consistent in all phases, if there isn't any phase, where just the 5% lower voltage causes way slower progress in some part compare to the rest.

Officially the lamps are rated to be dimmed down to less than 40% (so about 20W power for a 50W lamp) after the initial minimum 15minute warm up at full power. That means once the lamp reaches the unsaturated vapor state (so the arc voltage around the 100V mark), 20W should be sufficient to maintain it. So it means there is plenty of margin to keep the lamp in the correct state, once it reaches it, the only question when operating at lower power ballast is, whether and how reliably (so with how much margin) it will overcome all the warmup stages.