From the description of your takeoff it sure sounds like it's tail heavy. But on the other hand with a 15% CG located on the upper wing it can't possibly be so. It may be worth redoing the tail volume coefficient again. I think you may have gotten something messed up with a term or the MAC or some other issue.

From what you describe in the second takeoff with the ailerons making things worse is a classic pre stall adverse yaw problem. In those conditions you need to use rudder to help the ailerons. I also really think you are trying to lift off too fast or have too much up trim and that is causing it to lift off with the wing in a near stall condition. If ever there was a candidate for differential throw in the ailerons this model sounds like it.

How about trying a LOT of down trim so that the tail will lift up on the takeoff run to the point where you need to use a bit of back pressure to lift off? If you overdo it and suffer a prop strike then reduce the down trim and try again. The goal is to strongly avoid the nose up to a stall situation that seems to be the main problem. Be sure to rotate slowly with very light back pressure. The idea is to see the model's tail come up as it accelerates and then SLOOOWWWLY add back pressure until it's skimming and then add just a hair more to lift off. It's wise to let it accelerate at the skimming point for a moment before lifting it up. Once in the air but still in the ground effect to let it accellerate a bit more. It should then be safe to climb out at a shallow angle and then re trim the pitch from a safe altitude for level flight. A helper to reach over and add the trim for the first flight would be helpful since this one is obviously a handful and you may not want to take your thumbs off the sticks. A few clicks of right rudder trim for the first flights to help hold it away from the left spiral would not be a bad idea either.