Ah, the famous tip weight for takeoff. That's been discussed as well and it would not have a place in this model unless it turns out that it's out of balance laterally and you're just trying to correct it to put the lateral balance back onto the centerline. There's a thread about this article that came out in MA recently as well. It MAY, and I doubt even that, have a place in some trainers or scale models but it sure doesn't belong on any sort of aerobatic model.

I'd start by using something like the Robart incidence gauges to check for warps down to a high degree of accuracy and from there ensure that both ailerons are faired into the airfoil shape and not sticking up or down even slightly. Also check overall rigging of the wing to the fuselage and to the fin for even a hint of misalignment. Diagonal measurements from equal points out on the wing to the fuselage centerline back by the tail should be accurate to within 1/16 to 3/32 inch. And you can't trust the fuselage either. Set up some sort of rig to test the fuselage for straightness and that the fin is accurately mounted with a 1/2 degree of this tested and known good line.

If you read that other thread you'll find that there's a general consensus that radically tapered wings on our models are rather critical to a number of factors. Even small airfoil differences can cause what you're getting. But by far the worst thing is a heavy model and a strong taper ratio. It's just a bad combination that makes the model hyper sensitive to tip stalling. And if there's any sort of misalignment or side to side differences in the wing it'll encourage a stall and snap to one side or the other pretty consistently. About all you can do is try to find any minor differences, especially in the outer 1/2 of the wings, and "fix" them so both wings are identical.

Actually there MAY be something you can do to help this. On some of my gliders I've had good luck with adding turbulators to slightly delay the high drag mush that comes near the stall. They are also supposed to slightly increase the angle where the stall occurs. You MAY find you have some luck with adding turbulator strips on the outer third of the wing on the side that drops away consistently. On the other hand if it drops off either way about equally then try adding turbulators to both sides. Locating them to do the most good is a guessing game but I'd suggest a double thickness of 1/8 automotive trim striping tape at about the 8 to 10% chord point. If it doesn't help move this strip forward a little towards the 5% point.

Just be aware that a key element of this type of model is the ability to stall and snap violently on demand. To avoid it when NOT wanted it's up to the pilot to adjust to the model's needs. The pilots of the full sized airplanes have to do the same thing. Anyhow the big outcome of this is that the model may not snap the same upright as inverted or the same from one side to the other if you end up with the tubulator on the one side only. Adding more strips of different lengths may or may not fine tune the stalling charactarisitics but I just don't really know. At least it's something that's easy to play with and try different things.