The lowest voltage I have seen on a 4ft T12 tube is 71V, and that is from a 63W F48T12HO, presumably due to it's current being doubled. A close second is of course the 79V F34T12 energy saving version of F40T12, and that voltage drop is due to there being some krypton in there. Normal F40T12s are around 101V. So that got me thinking.

- Moving from .43A to .8A results in a ~30% decrease in lamp voltage.
- Moving from pure argon + mercury to a specific mixture of argon + krypton + mercury results in a ~22% decrease in lamp voltage.

What if we do both? Meaning have a lamp that is running at around .8A with a krypton+argon fill. I am sure it doesn't work this way, but if we use those same percentages we get 101*.7=71, 71*(79/101)=55.3V. So A 4ft T12 tube with a drop of ~55V. This is almost certainly low enough to be used with a choke on 120V. And what's more is that 55.3V*.8A=44.24VA, so it's wattage (and hopefully light output) is probably very similar to a normal F40T12, just higher current and lower voltage.

And what's more is that these specs are similar to the 37W F24T12HO (41V@.8A), so there might be compatibility there, no need for dedicated new ballast. If not, there is the possibility of using a 35W HPS choke ballast in a preheat configuration (55V@.83A).

Has anything like this been done before? Meaning two fluorescent lamps of identical physical dimensions and similar wattages that aren't electrically compatible? How possible would this lamp be? This seems like something American Fluorescent would've jumped on with their weird preheat-ish shop lights...

I would make a drawing out of this as usual, but there would probably be nothing visually different about it so that wouldn't be so productive.

Don't forget there's also the 48" VHO which is 110w/1500ma (my calculations gave 73v on that though)

There is the main question: Why would anyone do that?
The lowe the voltage, the grater percentage of that is the voltage drop associated with the electrodes, which generates heat but not light. So the lower the arc voltage the lower the resulting efficacy of the lamp.
So actually the opposite was happening: Making the lamps so the anode column drops as high voltage as possible, while not hitting some other limit.
One of the limits steering most of tradditional lamp sizes was the ability to operate with just a series choke ballast on the given mains. That is the reasom why you find so many lamps with arc voltage in the 50..60V ballpark (about maximum a 120V mains can support; F20T12, F15T8,...), or 90..120V (maximum for 230V mains (F40T12,...). Also when voltage boost is required, the cheapest and most efficient is when the ballast is doubling the mains as OCV (all windings use the same wire), which also leads to the 90..120V ballpark (and tubes like F40T12).
Another restriction, causing mainly the small lamps to deviate from the optimum above, is the intention to have common ballast for multiple lamp sizes (e.g. the miniature T5 family, where just the F8T5 is at the optimal 65V drop, the 6 and 4W are just designed to share the same ballast).
And the last reason to deviate from the optimum is the intention to make the lamp compatible with existing ballast, but lower the real power to save energy when the technology compensates by higher efficacy, like the F34T12.

So yes, it is possible to modify the lamp designs for lower arc voltage and higher currents, but there is no reason for doing that.