The corsair was flown on 22 volts and was seriously overpowered. But it wouldn't fly on only 3 cells as then it was way underpowered.

The "right" combination would probably have been 5 cells, but I don't have any 4400 mAh 2 cell packs, so 6 cells (3 x 2) was my only choice. Worked out well enough on the corsair, and you can clearly see in the video in the thread I linked you to above, the plane certainly had no trouble pulling it's weight around.

The trainer on the other hand is, if anything, underpowered on only 3 cells. Only reason I say this is because I have trouble taking off from the rough field I fly off of (same one as seen in the corsair video). The grass (especially now when everything is soggy after winter) provides a lot of resistance and the plane can't generate enough speed to get airborne when it's bouncing around. I'll be trying some larger props and bigger wheels to see if they help matters any.

If I take off from pavement the plane has no trouble getting airborne, and certainly, once it IS airborne, it has more than enough power to fly and climb at a pretty good rate, which is why I don't want to increase the voltage. The model flies nicely once airborne, just needs more power to take off from the (very poor) field I fly from.
4 cells would probably be perfect for the model all around, giving it plenty of power, but I haven't got any packs I can use to make up the 4 cells in voltage without dropping my pack capacity (my 4400 packs are all 3 cells, i've got 2200 2 cells, but they're only rated for 12C, and 2200 isn't very much for a large model like that). So i'll make do with the 3 cells.

Given current flight times and the fact the model, once airborne, will quite literally fly with almost no power, flights could probably be stretched to 15 minutes or more. Under full power, i'd guess around 8 to 10 minutes. Can't say for sure since it's a new model. Had to correct some CG issues after the maiden flight. It's all ready to go for it's next flight now, just need the weather to cooperate .

Obviously, what power will fly what plane varies very much on the design of the plane as well. The Tower Hobbies trainer happens to be very light and have a lot of wing area, so it doesn't take much to get it airborne (this is, actually, true of most trainers). But there's a lot of .40 sized models that are real pigs and require a lot of speed to get airborne, so it all depends. The corsair needed double the voltage to get airborne compared to the Tower Hobbies trainer, and they're nearly the same size (Corsair is heavier, though).

In my experience, it's much easier (and cheaper) to convert a plane to electric than a helicopter. Helicopters require quite a bit more power than a plane does to get airborne. With a plane, the model moving through the air and the wings airfoil does much of the lifting, so you can get away with quite a bit less power.
With a helicopter, ALL the power is being supplied by the motor and the comparatively small blades with their airfoil generating the lift. On top of this, the motor is always under near full power, so the entire electrical system is taxed more heavily.

This is why the same motor that made the corsair near ballistic in power is only adequate for the Hawk Pro on the same voltage. It's tricky stuff messing with electrics, a lot you need to take in and consider when deciding on what you need, but it can be fun, if a bit frustrating at times .