It is important to distinguish between imperfections that reflect design compromise or research failings, with the result that the product is inaccurate compared to the prototype, and imperfections that mean that the product is incomplete/damaged and/or does not function as intended. This thread records both, but it is the latter than concern me. Yes, an experienced modeller can correct some of the faults in the second category - it's sort of what we do. But test it this way. What if you had bought one of the faulty models for a sons/daughters birthday. Surely the minimum we expect from a product that is fit for purpose, that is, you can take it out of the box, plonk on the track and away it goes.
It doesn't matter whether the failure rate is 10%, 30% or 50% - the second category failures recorded in this thread are shocking, and we need to call that out.