Ok, the last few days I've been wondering if any type of High Output fluorescent bulb can be started and stay lit by using a typical preheat choke/ballast? 

I figured that the 2 foot version, a 24", 35w bulb would be best just for experiments. I don't have any HO bulbs, so I wouldn't know LOL
A F30/F40 preheat ballast, like a Robertson S40B would work, wouldn't it?

If if doesn't light properly, or stay lit, could it be a problem with the amount of current (mA) being supplied?

HO do use higher current to get higher power, you might be able to get the correct current by using larger choke (ie HID choke) which has lower inductance and made of thicker wire / core so is supposed to withstand higher current than fluorescent ballast, i dont know if high enough for your purpose

Or you could wire few ballasts in parallel. Assuming identical ballasts, the inductance will be divided by the number of ballasts, and the current carrying capacity multiplied

With preheaters is the problem usually with the current in the preheat phase: It should match the filament design, too small does not heat it up sufficiently - it's going more and more "blink happy" (as people use to call that behavior) and from some point (~60% and below) it would not start at all.
So as HO lamps are designed fr higher currents, regular "normal output" ballast would make the problem.
Too high current then evaporate he material from the electrodes, reducing their life.

But high current may be reduced by ~30% using a diode (1N4004 for up to 1A lamps) parallel to the filament - anode to the ballast side, cathode towards the starter side (it would effectively short out the filament, when acting as the anode).
But then for lag ballast (HX transformer,... - no capacitor in series with the lamp) this would ask for thermal protection, as if the filament broke, the setup would rectify and so drive the ballast into core saturation.
If the ballast uses series capacitor (e.g. MV CWA), the problem is not there, as the capacitor prevent the DC current component.

But how to determine, what current is needed by the filament?
First measure the filament resistance at cold.
Take (DC) variable power source (at least in the range 2.5..10V, capable to deliver at least double the rated lamp current), V-meter and the A-meter, connect it to one of the filaments and find the current, when the filament resistance is 4x the cold one (I guess you know, how to use the Ohm's law...) and measure the required current.
The correct preheat current is then 25..30% higher.
Then find a ballast, what deliver this current into short circuit. It's OCV should then be at least double the lamp arc voltage. Tolerance +/-20% should be still OK, but more you deviate, more problems to expect (see above).

Note, then with both end connected together (IS setup) the same lamp could handle up to twice the current then with the classic preheat circuit, so don't be surprised, then you get the optimum ballast current lower then the rated lamp current, the lamp rating correspond to the "perscribed" circuit, what is not what you are going to use.

Don't forget the old T17 bulbs and fixtures ARE HO !!

The thing, HO and VHO lamps (at least the ones in the USA) require the electrodes to be continuously heated for long life. That's why they are called rapid start, even though the ballasts typically start them almost instantly.

You can add inductor (another ballast) or resistor (incandescent) in series with the starter, to reduce starting current without changing normal work current

By today's standards, the F90T17s are actually VHO at 1500 Ma.  The HOs are 800 Ma.  So, to answer the question, YES, VHOs can be preheated!  Still working on the HOs though...

You can add inductor (another ballast) or resistor (incandescent) in series with the starter, to reduce starting current without changing normal work current

If the operating current would be too high, it would overheat electrodes bz itself - don§t forget, then it flow trough aprox. half of the filament length during normal operation, so if this would be too much, it would need to be reduced not only for the preheat, but for the normal operation as well (so the rac current would flow from both ends and not only from one side; this difference is there mainly on the cathode side, as on the anode the low impedance lead-in wire take the electrons, as it would have more positive potential)

But based on what Silverliner wrote, I guess here would be the opposite problem: The filament are not enough heated without external supply, what would then ask for higher operating current. But higher operating current may thermally overload the whole tube (yielding loss of efficacy due to improper mercury pressure, risk of premature failure of some seals, faster phosphor wear,...)