By spreading I meant thar if there is an underlying manufacturing defect it would be really difficult to predict where and when it would happen on memory. And if there is already a fault detected more might come later.

But I'm glad you documented how to do these types of workaround. It might be helpful with memory prices skyrocketing...