After learning about how fluorescent lights are wired up around the world, I have seen that in North American 2 lamp F20T12 preheat fixtures, I see that each tube almost always runs on their own independent reactor/choke ballast while most European 2 lamp F20T12 preheat fixtures run both tubes in series with their own starter circuits on a single lamp F40T12 reactor/choke ballast. After knowing that North American 2 lamp F20T12 preheat fixtures tend to have each tube running on their own reactor/choke ballast for minimizing the ballast losses associated with running both tubes in series on a single lamp F40T12 autotransformer preheat ballast, I wonder why it is the advantageous for European 2 lamp F20T12 preheat fixtures to have both tubes running in series on a single lamp F40T12 reactor/choke ballast?

Space, weight and cost.
This happens also on other wattage and lamp type

Ballast for single lamp would have about 220V across it, when a single ballast used for 2 lamps, the voltage across the ballast is around 195V.
So with 0.36A current, single ballast has to handle 79VA, so with Q=10 it means about 8W losses for each single lamp, so 16W losses for both lamps. With the single ballast for 2 lamps theballast has to handle 70VA, so with the same quality factor of 10 it means just 7W losses for both lamps, so less than half of losses and half of ballast cost.
The 230V mains is enough to support 2 F20 in series without boosting the voltage, so using a single ballast means you have less than half of the losses and ballast cost, hence the popularity...

I have run a pair of F20T12s in series on a single F40T12 preheat autotransformer ballast.  Worked quite well.

I have also been toying with the idea of running using a North American 2 lamp F40T12 lead lag preheat ballast as a 4 lamp lead lag F20T12 preheat ballast lately. In some cases, running 4 F20T12 tubes on some 2 lamp F40T12 ballasts would use 2 FS2 starters on one side and 1 FS4 starter for the pair of F20T12 tubes on the starting compensator side while in other cases, I would use 4 FS2 starters if the 2 lamp F40T12 ballast has 2 starter loops on the wiring diagram.

You can, indeed, run 2x F20 on tbe autotransformer F40 ballast, but the autotransformer is way heavier and lossier than the F20 choke, so it does not bring any ballast cost nor efficiency benefit over using just a simple choke for each lamp.
But if the question is to build something from parts you already have at home, it will endeed work well...

If the pair of F20T12 tubes was to be run in series on a single F40T12 reactor/choke ballast, would they have a shorter lifetime and output more lumens per tube when compared to operating each tube on an independent F20T12 reactor/choke ballast since the F20T12 pair would be overdriven at 430mA on a series circuit vs being operated at 380mA on independent reactor/choke ballasts?

No. The ballasts are not perfect current sources, their current tends to droop as the load voltage goes up (big extend is due to choke core or HX transformer shunt partial saturation). At least on the 230V series chokes it does so in a way the current matches the F20 rating. To me it really looks like the specs were intentionally set so a single ballast choke type matches both variants, a single F40, as well as a pair of F20's.

But not 100% sure if the US F40 HX preheat ballasts will behave the same way or not, because there it does not make much sense (cost-, as well as losses- wise) to use a single ballast for a series pair of tubes instead of way simpler, less lossy and cheaper individual chokes.

Do preheat lamps in series suffer from the issue of when one burns out the other wont light?

Do preheat lamps in series suffer from the issue of when one burns out the other wont light?

Yes, but usually both are of a similar wear, so when one fails, you just replace both, as the second one is very close to its EOL anyway. Plus the one failing usually causes quite excessive wear on the second one (it is practically repetitive starting, what the 2'nd lamp sees)

But indeed, reliability wise you have a situation "when single from a group fails, all fail" nature. But mostly the series pair uses to be just part of a 3- or 4-lamp fixtures (dropped ceiling "tiles",...), so still some redundancy make that up there.

I believe that if the bad lamp’s starter gets stuck quickly, then the bad starter will let the current bypass the EOL lamp and the other one will light, however it will be overdriven.

Not sure how i feel about those series systems lol. Prefer independant ballasts for preheat

They seem handy for something like a 2X2 troffer using 4X F20T12 lamps as you don't have to worry about installing 4 chokes and plus you get stroboscopic correction and power factor correction over individual chokes. You can wire up the chokes in a lead/lag configuration using a capacitor to get strobe and power factor correction but that's a rather unconventional wiring menthod. Plus I'm not sure you can fit that much (2 caps, 4 chokes, 2 starting comps) in a 2X2 troffer.   

Other than that, it seems like a bit of hassle in 120v areas to wire up a series preheat system (added ballast losses, more expensive ballast) compared to a individual choke per lamp. In 240v countries the equation is different as chokes are used for both 20w and 40w and my understanding is 40w chokes don't cost that much more than 20w chokes. 

Ballast costs and losses are somehow proportional to the sum of VA what all the windings have to deal with (multiply each windings voltage in an unloaded state with the rated current through that winding, sum the results from all windings up). Normally losses are 5..10% from that, you may use thicker wires to get losses down but you end up with more expensive and heavier unit or vice versa, but still using a configuration yielding lower VAs on the ballast means lower cost or losses or both.

For European 230V ballasts, both 40 and 20W are around 80VA chokes with about 6W losses (~7.5% from the total VA).
So operating each tube separately that means total 160VA chokes, with two in series just a single 80VA. So about half the cost and losses for the series combination, good enough incentive to live with the lower reliability (once there are more than a single pair of lamps illuminating the same space, as is most of the times the case for offices or so, the reliability is not any real issue anymore, even more when using cutout starters)


For the US 20W the ballast would be about 35VA, what means roughly 2.6W losses (assuming the same 7.5%VA design), for the F40 the 220V OCV autotransformer would have to handle 95VA so about 7W losses.

So operating two F20 on a single F40 ballast in the US would mean the need for a 95VA ballast with 7W losses, while two separate chokes would mean 70VA with 5.2W losses total. So both individual ballasts are for sure cheaper and waste less power than the single F40 ballast feeding both lamps in series, so it would be rather dumb to spend more money to get higher energy losses and lower reliability (because one failure takes down the second lamp too).
With fxF40 lead-lag the total VAS would be 173VA (assume 220V OCV and no compensator winding), so about 43VA per each F20 lamp. Still higher than the 35VA the single choke offers, so still it would be more expensive, more lossy and less reliable, so installing series lamp combinations does not make any sense at all.