B:
Don’t throw it out. In theory, a rigid wheelbase of > 3 wheels can’t make perfect contact. In practice, plenty of model steamers, including all of my own, have rigid wheelbases and work fine. If you’re lifting the wheel enough to clear the .025 RP25 flanges, ya gots track problems, buddy boy, big track problems.
I don’t have this engine, but I do have an MDC 2-6-0, which has an even longer driver separation, and it works fine. Anyway, the point is that the basic design isn’t inherently unsound, and works fine in practice, theory notwithstanding,so the problem certainly is fixable. In other words, don’t throw it out yet.
What I want to know is this: how is the lead truck designed on that? Is it mounted on a long leaf spring, pivoted behind the truck, or slotted, or what?
Do you have any more details? Does it derail more backing up, or going forward? Does it derail at switches, or is it always climbing one or the other rail at curves?
Meanwhile, I’ll leave you with one idea:

(The drawing of the derailment is a little crude, but the general idea is there. Basically, your truck will tend to go diagonal on the rail if the pivot arm is short.)
The extended-pivot arrangement, of course, can never be right for any but one radius. In practice, though, it works fine to fit it to your sharpest radius. The main thing is to keep the rail/wheel angle at a minimum. It doesn’t have to be zero. It never is, for any four wheel truck.
I’ll look at my MDC 2-6-0 and see if I can make it derail the same way, and report on the findings…